id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.7k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2101.01207 | Chloe He | Chloe He, Raksha Jain, J\'er\^ome Chambost, C\'eline Jacques, Cristina
Hickman | Semantic Video Segmentation for Intracytoplasmic Sperm Injection
Procedures | Accepted at the 'Medical Imaging meets NeurIPS Workshop' at the 34th
Conference on Neural Information Processing Systems | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present the first deep learning model for the analysis of intracytoplasmic
sperm injection (ICSI) procedures. Using a dataset of ICSI procedure videos, we
train a deep neural network to segment key objects in the videos achieving a
mean IoU of 0.962, and to localize the needle tip achieving a mean pixel error
of 3.793 pixels at 14 FPS on a single GPU. We further analyze the variation
between the dataset's human annotators and find the model's performance to be
comparable to human experts.
| [
{
"version": "v1",
"created": "Mon, 4 Jan 2021 19:33:12 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"He",
"Chloe",
""
],
[
"Jain",
"Raksha",
""
],
[
"Chambost",
"Jérôme",
""
],
[
"Jacques",
"Céline",
""
],
[
"Hickman",
"Cristina",
""
]
] | TITLE: Semantic Video Segmentation for Intracytoplasmic Sperm Injection
Procedures
ABSTRACT: We present the first deep learning model for the analysis of intracytoplasmic
sperm injection (ICSI) procedures. Using a dataset of ICSI procedure videos, we
train a deep neural network to segment key objects in the videos achieving a
mean IoU of 0.962, and to localize the needle tip achieving a mean pixel error
of 3.793 pixels at 14 FPS on a single GPU. We further analyze the variation
between the dataset's human annotators and find the model's performance to be
comparable to human experts.
|
2106.07955 | Matteo Rizzo | Matteo Rizzo, Cristina Conati, Daesik Jang, Hui Hu | Cascading Convolutional Temporal Colour Constancy | null | null | 10.1117/1.JEI.32.1.013049 | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Computational Colour Constancy (CCC) consists of estimating the colour of one
or more illuminants in a scene and using them to remove unwanted chromatic
distortions. Much research has focused on illuminant estimation for CCC on
single images, with few attempts of leveraging the temporal information
intrinsic in sequences of correlated images (e.g., the frames in a video), a
task known as Temporal Colour Constancy (TCC). The state-of-the-art for TCC is
TCCNet, a deep-learning architecture that uses a ConvLSTM for aggregating the
encodings produced by CNN submodules for each image in a sequence. We extend
this architecture with different models obtained by (i) substituting the TCCNet
submodules with C4, the state-of-the-art method for CCC targeting images; (ii)
adding a cascading strategy to perform an iterative improvement of the estimate
of the illuminant. We tested our models on the recently released TCC benchmark
and achieved results that surpass the state-of-the-art. Analyzing the impact of
the number of frames involved in illuminant estimation on performance, we show
that it is possible to reduce inference time by training the models on few
selected frames from the sequences while retaining comparable accuracy.
| [
{
"version": "v1",
"created": "Tue, 15 Jun 2021 08:17:30 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Rizzo",
"Matteo",
""
],
[
"Conati",
"Cristina",
""
],
[
"Jang",
"Daesik",
""
],
[
"Hu",
"Hui",
""
]
] | TITLE: Cascading Convolutional Temporal Colour Constancy
ABSTRACT: Computational Colour Constancy (CCC) consists of estimating the colour of one
or more illuminants in a scene and using them to remove unwanted chromatic
distortions. Much research has focused on illuminant estimation for CCC on
single images, with few attempts of leveraging the temporal information
intrinsic in sequences of correlated images (e.g., the frames in a video), a
task known as Temporal Colour Constancy (TCC). The state-of-the-art for TCC is
TCCNet, a deep-learning architecture that uses a ConvLSTM for aggregating the
encodings produced by CNN submodules for each image in a sequence. We extend
this architecture with different models obtained by (i) substituting the TCCNet
submodules with C4, the state-of-the-art method for CCC targeting images; (ii)
adding a cascading strategy to perform an iterative improvement of the estimate
of the illuminant. We tested our models on the recently released TCC benchmark
and achieved results that surpass the state-of-the-art. Analyzing the impact of
the number of frames involved in illuminant estimation on performance, we show
that it is possible to reduce inference time by training the models on few
selected frames from the sequences while retaining comparable accuracy.
|
2108.11250 | Dong Wu | Dong Wu, Manwen Liao, Weitian Zhang, Xinggang Wang, Xiang Bai, Wenqing
Cheng, Wenyu Liu | YOLOP: You Only Look Once for Panoptic Driving Perception | null | [J]. Machine Intelligence Research, 2022: 1-13 | 10.1007/s11633-022-1339-y | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | A panoptic driving perception system is an essential part of autonomous
driving. A high-precision and real-time perception system can assist the
vehicle in making the reasonable decision while driving. We present a panoptic
driving perception network (YOLOP) to perform traffic object detection,
drivable area segmentation and lane detection simultaneously. It is composed of
one encoder for feature extraction and three decoders to handle the specific
tasks. Our model performs extremely well on the challenging BDD100K dataset,
achieving state-of-the-art on all three tasks in terms of accuracy and speed.
Besides, we verify the effectiveness of our multi-task learning model for joint
training via ablative studies. To our best knowledge, this is the first work
that can process these three visual perception tasks simultaneously in
real-time on an embedded device Jetson TX2(23 FPS) and maintain excellent
accuracy. To facilitate further research, the source codes and pre-trained
models are released at https://github.com/hustvl/YOLOP.
| [
{
"version": "v1",
"created": "Wed, 25 Aug 2021 14:19:42 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Aug 2021 05:59:59 GMT"
},
{
"version": "v3",
"created": "Fri, 27 Aug 2021 06:31:48 GMT"
},
{
"version": "v4",
"created": "Mon, 30 Aug 2021 08:26:32 GMT"
},
{
"version": "v5",
"created": "Tue, 31 Aug 2021 08:38:29 GMT"
},
{
"version": "v6",
"created": "Fri, 11 Feb 2022 16:11:44 GMT"
},
{
"version": "v7",
"created": "Sat, 26 Mar 2022 15:39:42 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Wu",
"Dong",
""
],
[
"Liao",
"Manwen",
""
],
[
"Zhang",
"Weitian",
""
],
[
"Wang",
"Xinggang",
""
],
[
"Bai",
"Xiang",
""
],
[
"Cheng",
"Wenqing",
""
],
[
"Liu",
"Wenyu",
""
]
] | TITLE: YOLOP: You Only Look Once for Panoptic Driving Perception
ABSTRACT: A panoptic driving perception system is an essential part of autonomous
driving. A high-precision and real-time perception system can assist the
vehicle in making the reasonable decision while driving. We present a panoptic
driving perception network (YOLOP) to perform traffic object detection,
drivable area segmentation and lane detection simultaneously. It is composed of
one encoder for feature extraction and three decoders to handle the specific
tasks. Our model performs extremely well on the challenging BDD100K dataset,
achieving state-of-the-art on all three tasks in terms of accuracy and speed.
Besides, we verify the effectiveness of our multi-task learning model for joint
training via ablative studies. To our best knowledge, this is the first work
that can process these three visual perception tasks simultaneously in
real-time on an embedded device Jetson TX2(23 FPS) and maintain excellent
accuracy. To facilitate further research, the source codes and pre-trained
models are released at https://github.com/hustvl/YOLOP.
|
2109.04286 | Xiangyu Sun | Xiangyu Sun, Oliver Schulte, Guiliang Liu, Pascal Poupart | NTS-NOTEARS: Learning Nonparametric DBNs With Prior Knowledge | AISTATS 2023 | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe NTS-NOTEARS, a score-based structure learning method for
time-series data to learn dynamic Bayesian networks (DBNs) that captures
nonlinear, lagged (inter-slice) and instantaneous (intra-slice) relations among
variables. NTS-NOTEARS utilizes 1D convolutional neural networks (CNNs) to
model the dependence of child variables on their parents; 1D CNN is a neural
function approximation model well-suited for sequential data. DBN-CNN structure
learning is formulated as a continuous optimization problem with an acyclicity
constraint, following the NOTEARS DAG learning approach. We show how prior
knowledge of dependencies (e.g., forbidden and required edges) can be included
as additional optimization constraints. Empirical evaluation on simulated and
benchmark data show that NTS-NOTEARS achieves state-of-the-art DAG structure
quality compared to both parametric and nonparametric baseline methods, with
improvement in the range of 10-20% on the F1-score. We also evaluate
NTS-NOTEARS on complex real-world data acquired from professional ice hockey
games that contain a mixture of continuous and discrete variables. The code is
available online.
| [
{
"version": "v1",
"created": "Thu, 9 Sep 2021 14:08:09 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Oct 2022 20:54:58 GMT"
},
{
"version": "v3",
"created": "Wed, 1 Mar 2023 20:27:58 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Sun",
"Xiangyu",
""
],
[
"Schulte",
"Oliver",
""
],
[
"Liu",
"Guiliang",
""
],
[
"Poupart",
"Pascal",
""
]
] | TITLE: NTS-NOTEARS: Learning Nonparametric DBNs With Prior Knowledge
ABSTRACT: We describe NTS-NOTEARS, a score-based structure learning method for
time-series data to learn dynamic Bayesian networks (DBNs) that captures
nonlinear, lagged (inter-slice) and instantaneous (intra-slice) relations among
variables. NTS-NOTEARS utilizes 1D convolutional neural networks (CNNs) to
model the dependence of child variables on their parents; 1D CNN is a neural
function approximation model well-suited for sequential data. DBN-CNN structure
learning is formulated as a continuous optimization problem with an acyclicity
constraint, following the NOTEARS DAG learning approach. We show how prior
knowledge of dependencies (e.g., forbidden and required edges) can be included
as additional optimization constraints. Empirical evaluation on simulated and
benchmark data show that NTS-NOTEARS achieves state-of-the-art DAG structure
quality compared to both parametric and nonparametric baseline methods, with
improvement in the range of 10-20% on the F1-score. We also evaluate
NTS-NOTEARS on complex real-world data acquired from professional ice hockey
games that contain a mixture of continuous and discrete variables. The code is
available online.
|
2111.11138 | Hanan Salam | Hanan Salam | Distinguishing Engagement Facets: An Essential Component for AI-based
Interactive Healthcare | null | null | null | null | cs.HC cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Engagement in Human-Machine Interaction is the process by which entities
participating in the interaction establish, maintain, and end their perceived
connection. It is essential to monitor the engagement state of patients in
various AI-based interactive healthcare paradigms. This includes medical
conditions that alter social behavior such as Autism Spectrum Disorder (ASD) or
Attention-Deficit/Hyperactivity Disorder (ADHD). Engagement is a multi-faceted
construct which is composed of behavioral, emotional, and mental components.
Previous research has neglected this multi-faceted nature of engagement and
focused on the detection of engagement level or binary engagement label. In
this paper, a system is presented to distinguish these facets using contextual
and relational features. This can facilitate further fine-grained analysis.
Several machine learning classifiers including traditional and deep learning
models are compared for this task. An F-Score of 0.74 was obtained on a
balanced dataset of 22242 instances with neural network-based classification.
The proposed framework shall serve as a baseline for further research on
engagement facets recognition, and its integration is socially assistive
robotic applications.
| [
{
"version": "v1",
"created": "Mon, 22 Nov 2021 11:58:26 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 14:43:00 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Salam",
"Hanan",
""
]
] | TITLE: Distinguishing Engagement Facets: An Essential Component for AI-based
Interactive Healthcare
ABSTRACT: Engagement in Human-Machine Interaction is the process by which entities
participating in the interaction establish, maintain, and end their perceived
connection. It is essential to monitor the engagement state of patients in
various AI-based interactive healthcare paradigms. This includes medical
conditions that alter social behavior such as Autism Spectrum Disorder (ASD) or
Attention-Deficit/Hyperactivity Disorder (ADHD). Engagement is a multi-faceted
construct which is composed of behavioral, emotional, and mental components.
Previous research has neglected this multi-faceted nature of engagement and
focused on the detection of engagement level or binary engagement label. In
this paper, a system is presented to distinguish these facets using contextual
and relational features. This can facilitate further fine-grained analysis.
Several machine learning classifiers including traditional and deep learning
models are compared for this task. An F-Score of 0.74 was obtained on a
balanced dataset of 22242 instances with neural network-based classification.
The proposed framework shall serve as a baseline for further research on
engagement facets recognition, and its integration is socially assistive
robotic applications.
|
2111.13802 | Alasdair Tran | Alasdair Tran, Alexander Mathews, Lexing Xie, Cheng Soon Ong | Factorized Fourier Neural Operators | Published in The Eleventh International Conference on Learning
Representations (2023). Code is available at
https://github.com/alasdairtran/fourierflow | null | null | null | cs.LG cs.CE | http://creativecommons.org/licenses/by/4.0/ | We propose the Factorized Fourier Neural Operator (F-FNO), a learning-based
approach for simulating partial differential equations (PDEs). Starting from a
recently proposed Fourier representation of flow fields, the F-FNO bridges the
performance gap between pure machine learning approaches to that of the best
numerical or hybrid solvers. This is achieved with new representations -
separable spectral layers and improved residual connections - and a combination
of training strategies such as the Markov assumption, Gaussian noise, and
cosine learning rate decay. On several challenging benchmark PDEs on regular
grids, structured meshes, and point clouds, the F-FNO can scale to deeper
networks and outperform both the FNO and the geo-FNO, reducing the error by 83%
on the Navier-Stokes problem, 31% on the elasticity problem, 57% on the airfoil
flow problem, and 60% on the plastic forging problem. Compared to the
state-of-the-art pseudo-spectral method, the F-FNO can take a step size that is
an order of magnitude larger in time and achieve an order of magnitude speedup
to produce the same solution quality.
| [
{
"version": "v1",
"created": "Sat, 27 Nov 2021 03:34:13 GMT"
},
{
"version": "v2",
"created": "Tue, 30 Nov 2021 02:04:46 GMT"
},
{
"version": "v3",
"created": "Thu, 15 Dec 2022 10:54:22 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Mar 2023 12:40:23 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Tran",
"Alasdair",
""
],
[
"Mathews",
"Alexander",
""
],
[
"Xie",
"Lexing",
""
],
[
"Ong",
"Cheng Soon",
""
]
] | TITLE: Factorized Fourier Neural Operators
ABSTRACT: We propose the Factorized Fourier Neural Operator (F-FNO), a learning-based
approach for simulating partial differential equations (PDEs). Starting from a
recently proposed Fourier representation of flow fields, the F-FNO bridges the
performance gap between pure machine learning approaches to that of the best
numerical or hybrid solvers. This is achieved with new representations -
separable spectral layers and improved residual connections - and a combination
of training strategies such as the Markov assumption, Gaussian noise, and
cosine learning rate decay. On several challenging benchmark PDEs on regular
grids, structured meshes, and point clouds, the F-FNO can scale to deeper
networks and outperform both the FNO and the geo-FNO, reducing the error by 83%
on the Navier-Stokes problem, 31% on the elasticity problem, 57% on the airfoil
flow problem, and 60% on the plastic forging problem. Compared to the
state-of-the-art pseudo-spectral method, the F-FNO can take a step size that is
an order of magnitude larger in time and achieve an order of magnitude speedup
to produce the same solution quality.
|
2112.11396 | Jonathan Cardoso-Silva | Caterina De Bacco, Martina Contisciani, Jonathan Cardoso-Silva,
Hadiseh Safdari, Diego Baptista, Gabriela L. Borges, Tracy Sweet,
Jean-Gabriel Young, Jeremy Koster, Cody T. Ross, Richard McElreath, Daniel
Redhead, Eleanor A. Power | Latent Network Models to Account for Noisy, Multiply-Reported Social
Network Data | null | null | 10.1093/jrsssa/qnac004 | null | cs.SI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Social network data are often constructed by incorporating reports from
multiple individuals. However, it is not obvious how to reconcile discordant
responses from individuals. There may be particular risks with
multiply-reported data if people's responses reflect normative expectations --
such as an expectation of balanced, reciprocal relationships. Here, we propose
a probabilistic model that incorporates ties reported by multiple individuals
to estimate the unobserved network structure. In addition to estimating a
parameter for each reporter that is related to their tendency of over- or
under-reporting relationships, the model explicitly incorporates a term for
``mutuality,'' the tendency to report ties in both directions involving the
same alter. Our model's algorithmic implementation is based on variational
inference, which makes it efficient and scalable to large systems. We apply our
model to data from 75 Indian villages collected with a name-generator design,
and a Nicaraguan community collected with a roster-based design. We observe
strong evidence of ``mutuality'' in both datasets, and find that this value
varies by relationship type. Consequently, our model estimates networks with
reciprocity values that are substantially different than those resulting from
standard deterministic aggregation approaches, demonstrating the need to
consider such issues when gathering, constructing, and analysing survey-based
network data.
| [
{
"version": "v1",
"created": "Tue, 21 Dec 2021 17:55:07 GMT"
},
{
"version": "v2",
"created": "Mon, 12 Dec 2022 15:57:18 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"De Bacco",
"Caterina",
""
],
[
"Contisciani",
"Martina",
""
],
[
"Cardoso-Silva",
"Jonathan",
""
],
[
"Safdari",
"Hadiseh",
""
],
[
"Baptista",
"Diego",
""
],
[
"Borges",
"Gabriela L.",
""
],
[
"Sweet",
"Tracy",
""
],
[
"Young",
"Jean-Gabriel",
""
],
[
"Koster",
"Jeremy",
""
],
[
"Ross",
"Cody T.",
""
],
[
"McElreath",
"Richard",
""
],
[
"Redhead",
"Daniel",
""
],
[
"Power",
"Eleanor A.",
""
]
] | TITLE: Latent Network Models to Account for Noisy, Multiply-Reported Social
Network Data
ABSTRACT: Social network data are often constructed by incorporating reports from
multiple individuals. However, it is not obvious how to reconcile discordant
responses from individuals. There may be particular risks with
multiply-reported data if people's responses reflect normative expectations --
such as an expectation of balanced, reciprocal relationships. Here, we propose
a probabilistic model that incorporates ties reported by multiple individuals
to estimate the unobserved network structure. In addition to estimating a
parameter for each reporter that is related to their tendency of over- or
under-reporting relationships, the model explicitly incorporates a term for
``mutuality,'' the tendency to report ties in both directions involving the
same alter. Our model's algorithmic implementation is based on variational
inference, which makes it efficient and scalable to large systems. We apply our
model to data from 75 Indian villages collected with a name-generator design,
and a Nicaraguan community collected with a roster-based design. We observe
strong evidence of ``mutuality'' in both datasets, and find that this value
varies by relationship type. Consequently, our model estimates networks with
reciprocity values that are substantially different than those resulting from
standard deterministic aggregation approaches, demonstrating the need to
consider such issues when gathering, constructing, and analysing survey-based
network data.
|
2201.11945 | Lingxiao Li | Lingxiao Li, Noam Aigerman, Vladimir G. Kim, Jiajin Li, Kristjan
Greenewald, Mikhail Yurochkin, Justin Solomon | Learning Proximal Operators to Discover Multiple Optima | null | null | null | null | cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | Finding multiple solutions of non-convex optimization problems is a
ubiquitous yet challenging task. Most past algorithms either apply
single-solution optimization methods from multiple random initial guesses or
search in the vicinity of found solutions using ad hoc heuristics. We present
an end-to-end method to learn the proximal operator of a family of training
problems so that multiple local minima can be quickly obtained from initial
guesses by iterating the learned operator, emulating the proximal-point
algorithm that has fast convergence. The learned proximal operator can be
further generalized to recover multiple optima for unseen problems at test
time, enabling applications such as object detection. The key ingredient in our
formulation is a proximal regularization term, which elevates the convexity of
our training loss: by applying recent theoretical results, we show that for
weakly-convex objectives with Lipschitz gradients, training of the proximal
operator converges globally with a practical degree of over-parameterization.
We further present an exhaustive benchmark for multi-solution optimization to
demonstrate the effectiveness of our method.
| [
{
"version": "v1",
"created": "Fri, 28 Jan 2022 05:53:28 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Oct 2022 17:20:47 GMT"
},
{
"version": "v3",
"created": "Wed, 1 Mar 2023 20:52:31 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Li",
"Lingxiao",
""
],
[
"Aigerman",
"Noam",
""
],
[
"Kim",
"Vladimir G.",
""
],
[
"Li",
"Jiajin",
""
],
[
"Greenewald",
"Kristjan",
""
],
[
"Yurochkin",
"Mikhail",
""
],
[
"Solomon",
"Justin",
""
]
] | TITLE: Learning Proximal Operators to Discover Multiple Optima
ABSTRACT: Finding multiple solutions of non-convex optimization problems is a
ubiquitous yet challenging task. Most past algorithms either apply
single-solution optimization methods from multiple random initial guesses or
search in the vicinity of found solutions using ad hoc heuristics. We present
an end-to-end method to learn the proximal operator of a family of training
problems so that multiple local minima can be quickly obtained from initial
guesses by iterating the learned operator, emulating the proximal-point
algorithm that has fast convergence. The learned proximal operator can be
further generalized to recover multiple optima for unseen problems at test
time, enabling applications such as object detection. The key ingredient in our
formulation is a proximal regularization term, which elevates the convexity of
our training loss: by applying recent theoretical results, we show that for
weakly-convex objectives with Lipschitz gradients, training of the proximal
operator converges globally with a practical degree of over-parameterization.
We further present an exhaustive benchmark for multi-solution optimization to
demonstrate the effectiveness of our method.
|
2203.14291 | Ge-Peng Ji | Ge-Peng Ji, Guobao Xiao, Yu-Cheng Chou, Deng-Ping Fan, Kai Zhao, Geng
Chen, and Luc Van Gool | Video Polyp Segmentation: A Deep Learning Perspective | Accepted by Machine Intelligence Research 2022 (Project Page:
https://github.com/GewelsJI/VPS) | Machine Intelligence Research, vol. 19, no. 6, pp.531-549, 2022 | 10.1007/s11633-022-1371-y | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present the first comprehensive video polyp segmentation (VPS) study in
the deep learning era. Over the years, developments in VPS are not moving
forward with ease due to the lack of large-scale fine-grained segmentation
annotations. To address this issue, we first introduce a high-quality
frame-by-frame annotated VPS dataset, named SUN-SEG, which contains 158,690
colonoscopy frames from the well-known SUN-database. We provide additional
annotations with diverse types, i.e., attribute, object mask, boundary,
scribble, and polygon. Second, we design a simple but efficient baseline,
dubbed PNS+, consisting of a global encoder, a local encoder, and normalized
self-attention (NS) blocks. The global and local encoders receive an anchor
frame and multiple successive frames to extract long-term and short-term
spatial-temporal representations, which are then progressively updated by two
NS blocks. Extensive experiments show that PNS+ achieves the best performance
and real-time inference speed (170fps), making it a promising solution for the
VPS task. Third, we extensively evaluate 13 representative polyp/object
segmentation models on our SUN-SEG dataset and provide attribute-based
comparisons. Finally, we discuss several open issues and suggest possible
research directions for the VPS community.
| [
{
"version": "v1",
"created": "Sun, 27 Mar 2022 12:40:10 GMT"
},
{
"version": "v2",
"created": "Sat, 2 Jul 2022 23:08:48 GMT"
},
{
"version": "v3",
"created": "Wed, 31 Aug 2022 11:02:30 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Ji",
"Ge-Peng",
""
],
[
"Xiao",
"Guobao",
""
],
[
"Chou",
"Yu-Cheng",
""
],
[
"Fan",
"Deng-Ping",
""
],
[
"Zhao",
"Kai",
""
],
[
"Chen",
"Geng",
""
],
[
"Van Gool",
"Luc",
""
]
] | TITLE: Video Polyp Segmentation: A Deep Learning Perspective
ABSTRACT: We present the first comprehensive video polyp segmentation (VPS) study in
the deep learning era. Over the years, developments in VPS are not moving
forward with ease due to the lack of large-scale fine-grained segmentation
annotations. To address this issue, we first introduce a high-quality
frame-by-frame annotated VPS dataset, named SUN-SEG, which contains 158,690
colonoscopy frames from the well-known SUN-database. We provide additional
annotations with diverse types, i.e., attribute, object mask, boundary,
scribble, and polygon. Second, we design a simple but efficient baseline,
dubbed PNS+, consisting of a global encoder, a local encoder, and normalized
self-attention (NS) blocks. The global and local encoders receive an anchor
frame and multiple successive frames to extract long-term and short-term
spatial-temporal representations, which are then progressively updated by two
NS blocks. Extensive experiments show that PNS+ achieves the best performance
and real-time inference speed (170fps), making it a promising solution for the
VPS task. Third, we extensively evaluate 13 representative polyp/object
segmentation models on our SUN-SEG dataset and provide attribute-based
comparisons. Finally, we discuss several open issues and suggest possible
research directions for the VPS community.
|
2204.07946 | Ginam Kim | Ginam Kim, Hyunsung Kim, Joseph Kihoon Kim, Sung-Sik Cho, Yeong-Hun
Park, Suk-Ju Kang | Integrated In-vehicle Monitoring System Using 3D Human Pose Estimation
and Seat Belt Segmentation | AAAI 2022 workshop AI for Transportation accepted | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, along with interest in autonomous vehicles, the importance of
monitoring systems for both drivers and passengers inside vehicles has been
increasing. This paper proposes a novel in-vehicle monitoring system the
combines 3D pose estimation, seat-belt segmentation, and seat-belt status
classification networks. Our system outputs various information necessary for
monitoring by accurately considering the data characteristics of the in-vehicle
environment. Specifically, the proposed 3D pose estimation directly estimates
the absolute coordinates of keypoints for a driver and passengers, and the
proposed seat-belt segmentation is implemented by applying a structure based on
the feature pyramid. In addition, we propose a classification task to
distinguish between normal and abnormal states of wearing a seat belt using
results that combine 3D pose estimation with seat-belt segmentation. These
tasks can be learned simultaneously and operate in real-time. Our method was
evaluated on a private dataset we newly created and annotated. The experimental
results show that our method has significantly high performance that can be
applied directly to real in-vehicle monitoring systems.
| [
{
"version": "v1",
"created": "Sun, 17 Apr 2022 07:20:50 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 01:42:42 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kim",
"Ginam",
""
],
[
"Kim",
"Hyunsung",
""
],
[
"Kim",
"Joseph Kihoon",
""
],
[
"Cho",
"Sung-Sik",
""
],
[
"Park",
"Yeong-Hun",
""
],
[
"Kang",
"Suk-Ju",
""
]
] | TITLE: Integrated In-vehicle Monitoring System Using 3D Human Pose Estimation
and Seat Belt Segmentation
ABSTRACT: Recently, along with interest in autonomous vehicles, the importance of
monitoring systems for both drivers and passengers inside vehicles has been
increasing. This paper proposes a novel in-vehicle monitoring system the
combines 3D pose estimation, seat-belt segmentation, and seat-belt status
classification networks. Our system outputs various information necessary for
monitoring by accurately considering the data characteristics of the in-vehicle
environment. Specifically, the proposed 3D pose estimation directly estimates
the absolute coordinates of keypoints for a driver and passengers, and the
proposed seat-belt segmentation is implemented by applying a structure based on
the feature pyramid. In addition, we propose a classification task to
distinguish between normal and abnormal states of wearing a seat belt using
results that combine 3D pose estimation with seat-belt segmentation. These
tasks can be learned simultaneously and operate in real-time. Our method was
evaluated on a private dataset we newly created and annotated. The experimental
results show that our method has significantly high performance that can be
applied directly to real in-vehicle monitoring systems.
|
2205.03307 | Jiaqi Gao | Jiaqi Gao, Jingqi Li, Hongming Shan, Yanyun Qu, James Z. Wang, Fei-Yue
Wang, Junping Zhang | Forget Less, Count Better: A Domain-Incremental Self-Distillation
Learning Benchmark for Lifelong Crowd Counting | This paper has been accepted by Frontiers of Information Technology &
Electronic Engineering | Front Inform Technol Electron Eng 24, 187-202 (2023) | 10.1631/FITEE.2200380 | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Crowd counting has important applications in public safety and pandemic
control. A robust and practical crowd counting system has to be capable of
continuously learning with the new incoming domain data in real-world scenarios
instead of fitting one domain only. Off-the-shelf methods have some drawbacks
when handling multiple domains: (1) the models will achieve limited performance
(even drop dramatically) among old domains after training images from new
domains due to the discrepancies of intrinsic data distributions from various
domains, which is called catastrophic forgetting; (2) the well-trained model in
a specific domain achieves imperfect performance among other unseen domains
because of the domain shift; and (3) it leads to linearly increasing storage
overhead, either mixing all the data for training or simply training dozens of
separate models for different domains when new ones are available. To overcome
these issues, we investigated a new crowd counting task in the incremental
domains training setting called Lifelong Crowd Counting. Its goal is to
alleviate the catastrophic forgetting and improve the generalization ability
using a single model updated by the incremental domains. Specifically, we
propose a self-distillation learning framework as a benchmark (Forget Less,
Count Better, or FLCB) for lifelong crowd counting, which helps the model
sustainably leverage previous meaningful knowledge for better crowd counting to
mitigate the forgetting when the new data arrive. In addition, a new
quantitative metric, normalized backward transfer (nBwT), is developed to
evaluate the forgetting degree of the model in the lifelong learning process.
Extensive experimental results demonstrate the superiority of our proposed
benchmark in achieving a low catastrophic forgetting degree and strong
generalization ability.
| [
{
"version": "v1",
"created": "Fri, 6 May 2022 15:37:56 GMT"
},
{
"version": "v2",
"created": "Sun, 29 Jan 2023 06:57:19 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Gao",
"Jiaqi",
""
],
[
"Li",
"Jingqi",
""
],
[
"Shan",
"Hongming",
""
],
[
"Qu",
"Yanyun",
""
],
[
"Wang",
"James Z.",
""
],
[
"Wang",
"Fei-Yue",
""
],
[
"Zhang",
"Junping",
""
]
] | TITLE: Forget Less, Count Better: A Domain-Incremental Self-Distillation
Learning Benchmark for Lifelong Crowd Counting
ABSTRACT: Crowd counting has important applications in public safety and pandemic
control. A robust and practical crowd counting system has to be capable of
continuously learning with the new incoming domain data in real-world scenarios
instead of fitting one domain only. Off-the-shelf methods have some drawbacks
when handling multiple domains: (1) the models will achieve limited performance
(even drop dramatically) among old domains after training images from new
domains due to the discrepancies of intrinsic data distributions from various
domains, which is called catastrophic forgetting; (2) the well-trained model in
a specific domain achieves imperfect performance among other unseen domains
because of the domain shift; and (3) it leads to linearly increasing storage
overhead, either mixing all the data for training or simply training dozens of
separate models for different domains when new ones are available. To overcome
these issues, we investigated a new crowd counting task in the incremental
domains training setting called Lifelong Crowd Counting. Its goal is to
alleviate the catastrophic forgetting and improve the generalization ability
using a single model updated by the incremental domains. Specifically, we
propose a self-distillation learning framework as a benchmark (Forget Less,
Count Better, or FLCB) for lifelong crowd counting, which helps the model
sustainably leverage previous meaningful knowledge for better crowd counting to
mitigate the forgetting when the new data arrive. In addition, a new
quantitative metric, normalized backward transfer (nBwT), is developed to
evaluate the forgetting degree of the model in the lifelong learning process.
Extensive experimental results demonstrate the superiority of our proposed
benchmark in achieving a low catastrophic forgetting degree and strong
generalization ability.
|
2205.12502 | Gi-Cheon Kang | Gi-Cheon Kang, Sungdong Kim, Jin-Hwa Kim, Donghyun Kwak, Byoung-Tak
Zhang | The Dialog Must Go On: Improving Visual Dialog via Generative
Self-Training | CVPR 2023 | null | null | null | cs.CV cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Visual dialog (VisDial) is a task of answering a sequence of questions
grounded in an image, using the dialog history as context. Prior work has
trained the dialog agents solely on VisDial data via supervised learning or
leveraged pre-training on related vision-and-language datasets. This paper
presents a semi-supervised learning approach for visually-grounded dialog,
called Generative Self-Training (GST), to leverage unlabeled images on the Web.
Specifically, GST first retrieves in-domain images through out-of-distribution
detection and generates synthetic dialogs regarding the images via multimodal
conditional text generation. GST then trains a dialog agent on the synthetic
and the original VisDial data. As a result, GST scales the amount of training
data up to an order of magnitude that of VisDial (1.2M to 12.9M QA data). For
robust training of the synthetic dialogs, we also propose perplexity-based data
selection and multimodal consistency regularization. Evaluation on VisDial v1.0
and v0.9 datasets shows that GST achieves new state-of-the-art results on both
datasets. We further observe the robustness of GST against both visual and
textual adversarial attacks. Finally, GST yields strong performance gains in
the low-data regime. Code is available at
https://github.com/gicheonkang/gst-visdial.
| [
{
"version": "v1",
"created": "Wed, 25 May 2022 05:40:00 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 12:33:10 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kang",
"Gi-Cheon",
""
],
[
"Kim",
"Sungdong",
""
],
[
"Kim",
"Jin-Hwa",
""
],
[
"Kwak",
"Donghyun",
""
],
[
"Zhang",
"Byoung-Tak",
""
]
] | TITLE: The Dialog Must Go On: Improving Visual Dialog via Generative
Self-Training
ABSTRACT: Visual dialog (VisDial) is a task of answering a sequence of questions
grounded in an image, using the dialog history as context. Prior work has
trained the dialog agents solely on VisDial data via supervised learning or
leveraged pre-training on related vision-and-language datasets. This paper
presents a semi-supervised learning approach for visually-grounded dialog,
called Generative Self-Training (GST), to leverage unlabeled images on the Web.
Specifically, GST first retrieves in-domain images through out-of-distribution
detection and generates synthetic dialogs regarding the images via multimodal
conditional text generation. GST then trains a dialog agent on the synthetic
and the original VisDial data. As a result, GST scales the amount of training
data up to an order of magnitude that of VisDial (1.2M to 12.9M QA data). For
robust training of the synthetic dialogs, we also propose perplexity-based data
selection and multimodal consistency regularization. Evaluation on VisDial v1.0
and v0.9 datasets shows that GST achieves new state-of-the-art results on both
datasets. We further observe the robustness of GST against both visual and
textual adversarial attacks. Finally, GST yields strong performance gains in
the low-data regime. Code is available at
https://github.com/gicheonkang/gst-visdial.
|
2205.13192 | Joshua Pinskier | Thomas Lowe and Joshua Pinskier | Tree Reconstruction using Topology Optimisation | The datasets generated and used in the current study are available in
the Tree Reconstructions from Pointclouds Scanned in Pullenvale QLD
repository, https://doi.org/10.25919/yt2m-9373 | https://www.mdpi.com/2072-4292/15/1/172 | 10.3390/rs15010172 | null | cs.CV cs.GR cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Generating accurate digital tree models from scanned environments is
invaluable for forestry, agriculture, and other outdoor industries in tasks
such as identifying biomass, fall hazards and traversability, as well as
digital applications such as animation and gaming. Existing methods for tree
reconstruction rely on feature identification (trunk, crown, etc) to
heuristically segment a forest into individual trees and generate a branch
structure graph, limiting their application to sparse trees and uniform
forests. However, the natural world is a messy place in which trees present
with significant heterogeneity and are frequently encroached upon by the
surrounding environment. We present a general method for extracting the branch
structure of trees from point cloud data, which estimates the structure of
trees by adapting the methods of structural topology optimisation to find the
optimal material distribution to support wind-loading. We present the results
of this optimisation over a wide variety of scans, and discuss the benefits and
drawbacks of this novel approach to tree structure reconstruction. Despite the
high variability of datasets containing trees, and the high rate of occlusions,
our method generates detailed and accurate tree structures in most cases.
| [
{
"version": "v1",
"created": "Thu, 26 May 2022 07:08:32 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 00:27:11 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lowe",
"Thomas",
""
],
[
"Pinskier",
"Joshua",
""
]
] | TITLE: Tree Reconstruction using Topology Optimisation
ABSTRACT: Generating accurate digital tree models from scanned environments is
invaluable for forestry, agriculture, and other outdoor industries in tasks
such as identifying biomass, fall hazards and traversability, as well as
digital applications such as animation and gaming. Existing methods for tree
reconstruction rely on feature identification (trunk, crown, etc) to
heuristically segment a forest into individual trees and generate a branch
structure graph, limiting their application to sparse trees and uniform
forests. However, the natural world is a messy place in which trees present
with significant heterogeneity and are frequently encroached upon by the
surrounding environment. We present a general method for extracting the branch
structure of trees from point cloud data, which estimates the structure of
trees by adapting the methods of structural topology optimisation to find the
optimal material distribution to support wind-loading. We present the results
of this optimisation over a wide variety of scans, and discuss the benefits and
drawbacks of this novel approach to tree structure reconstruction. Despite the
high variability of datasets containing trees, and the high rate of occlusions,
our method generates detailed and accurate tree structures in most cases.
|
2205.14083 | Jiawei Du | Jiawei Du, Daquan Zhou, Jiashi Feng, Vincent Y. F. Tan, Joey Tianyi
Zhou | Sharpness-Aware Training for Free | null | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern deep neural networks (DNNs) have achieved state-of-the-art
performances but are typically over-parameterized. The over-parameterization
may result in undesirably large generalization error in the absence of other
customized training strategies. Recently, a line of research under the name of
Sharpness-Aware Minimization (SAM) has shown that minimizing a sharpness
measure, which reflects the geometry of the loss landscape, can significantly
reduce the generalization error. However, SAM-like methods incur a two-fold
computational overhead of the given base optimizer (e.g. SGD) for approximating
the sharpness measure. In this paper, we propose Sharpness-Aware Training for
Free, or SAF, which mitigates the sharp landscape at almost zero additional
computational cost over the base optimizer. Intuitively, SAF achieves this by
avoiding sudden drops in the loss in the sharp local minima throughout the
trajectory of the updates of the weights. Specifically, we suggest a novel
trajectory loss, based on the KL-divergence between the outputs of DNNs with
the current weights and past weights, as a replacement of the SAM's sharpness
measure. This loss captures the rate of change of the training loss along the
model's update trajectory. By minimizing it, SAF ensures the convergence to a
flat minimum with improved generalization capabilities. Extensive empirical
results show that SAF minimizes the sharpness in the same way that SAM does,
yielding better results on the ImageNet dataset with essentially the same
computational cost as the base optimizer.
| [
{
"version": "v1",
"created": "Fri, 27 May 2022 16:32:43 GMT"
},
{
"version": "v2",
"created": "Mon, 30 May 2022 02:13:20 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 06:39:34 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Du",
"Jiawei",
""
],
[
"Zhou",
"Daquan",
""
],
[
"Feng",
"Jiashi",
""
],
[
"Tan",
"Vincent Y. F.",
""
],
[
"Zhou",
"Joey Tianyi",
""
]
] | TITLE: Sharpness-Aware Training for Free
ABSTRACT: Modern deep neural networks (DNNs) have achieved state-of-the-art
performances but are typically over-parameterized. The over-parameterization
may result in undesirably large generalization error in the absence of other
customized training strategies. Recently, a line of research under the name of
Sharpness-Aware Minimization (SAM) has shown that minimizing a sharpness
measure, which reflects the geometry of the loss landscape, can significantly
reduce the generalization error. However, SAM-like methods incur a two-fold
computational overhead of the given base optimizer (e.g. SGD) for approximating
the sharpness measure. In this paper, we propose Sharpness-Aware Training for
Free, or SAF, which mitigates the sharp landscape at almost zero additional
computational cost over the base optimizer. Intuitively, SAF achieves this by
avoiding sudden drops in the loss in the sharp local minima throughout the
trajectory of the updates of the weights. Specifically, we suggest a novel
trajectory loss, based on the KL-divergence between the outputs of DNNs with
the current weights and past weights, as a replacement of the SAM's sharpness
measure. This loss captures the rate of change of the training loss along the
model's update trajectory. By minimizing it, SAF ensures the convergence to a
flat minimum with improved generalization capabilities. Extensive empirical
results show that SAF minimizes the sharpness in the same way that SAM does,
yielding better results on the ImageNet dataset with essentially the same
computational cost as the base optimizer.
|
2205.14589 | Tao Huang | Tao Huang, Yuan Zhang, Shan You, Fei Wang, Chen Qian, Jian Cao, Chang
Xu | Masked Distillation with Receptive Tokens | ICLR 2023 | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Distilling from the feature maps can be fairly effective for dense prediction
tasks since both the feature discriminability and localization priors can be
well transferred. However, not every pixel contributes equally to the
performance, and a good student should learn from what really matters to the
teacher. In this paper, we introduce a learnable embedding dubbed receptive
token to localize those pixels of interests (PoIs) in the feature map, with a
distillation mask generated via pixel-wise attention. Then the distillation
will be performed on the mask via pixel-wise reconstruction. In this way, a
distillation mask actually indicates a pattern of pixel dependencies within
feature maps of teacher. We thus adopt multiple receptive tokens to investigate
more sophisticated and informative pixel dependencies to further enhance the
distillation. To obtain a group of masks, the receptive tokens are learned via
the regular task loss but with teacher fixed, and we also leverage a Dice loss
to enrich the diversity of learned masks. Our method dubbed MasKD is simple and
practical, and needs no priors of tasks in application. Experiments show that
our MasKD can achieve state-of-the-art performance consistently on object
detection and semantic segmentation benchmarks. Code is available at:
https://github.com/hunto/MasKD .
| [
{
"version": "v1",
"created": "Sun, 29 May 2022 07:32:00 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 14:10:40 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Huang",
"Tao",
""
],
[
"Zhang",
"Yuan",
""
],
[
"You",
"Shan",
""
],
[
"Wang",
"Fei",
""
],
[
"Qian",
"Chen",
""
],
[
"Cao",
"Jian",
""
],
[
"Xu",
"Chang",
""
]
] | TITLE: Masked Distillation with Receptive Tokens
ABSTRACT: Distilling from the feature maps can be fairly effective for dense prediction
tasks since both the feature discriminability and localization priors can be
well transferred. However, not every pixel contributes equally to the
performance, and a good student should learn from what really matters to the
teacher. In this paper, we introduce a learnable embedding dubbed receptive
token to localize those pixels of interests (PoIs) in the feature map, with a
distillation mask generated via pixel-wise attention. Then the distillation
will be performed on the mask via pixel-wise reconstruction. In this way, a
distillation mask actually indicates a pattern of pixel dependencies within
feature maps of teacher. We thus adopt multiple receptive tokens to investigate
more sophisticated and informative pixel dependencies to further enhance the
distillation. To obtain a group of masks, the receptive tokens are learned via
the regular task loss but with teacher fixed, and we also leverage a Dice loss
to enrich the diversity of learned masks. Our method dubbed MasKD is simple and
practical, and needs no priors of tasks in application. Experiments show that
our MasKD can achieve state-of-the-art performance consistently on object
detection and semantic segmentation benchmarks. Code is available at:
https://github.com/hunto/MasKD .
|
2206.04624 | Wei Ping | Nayeon Lee, Wei Ping, Peng Xu, Mostofa Patwary, Pascale Fung, Mohammad
Shoeybi, Bryan Catanzaro | Factuality Enhanced Language Models for Open-Ended Text Generation | NeurIPS 2022 | null | null | null | cs.CL cs.AI cs.CY cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pretrained language models (LMs) are susceptible to generate text with
nonfactual information. In this work, we measure and improve the factual
accuracy of large-scale LMs for open-ended text generation. We design the
FactualityPrompts test set and metrics to measure the factuality of LM
generations. Based on that, we study the factual accuracy of LMs with parameter
sizes ranging from 126M to 530B. Interestingly, we find that larger LMs are
more factual than smaller ones, although a previous study suggests that larger
LMs can be less truthful in terms of misconceptions. In addition, popular
sampling algorithms (e.g., top-p) in open-ended text generation can harm the
factuality due to the ''uniform randomness'' introduced at every sampling step.
We propose the factual-nucleus sampling algorithm that dynamically adapts the
randomness to improve the factuality of generation while maintaining quality.
Furthermore, we analyze the inefficiencies of the standard training method in
learning correct associations between entities from factual text corpus (e.g.,
Wikipedia). We propose a factuality-enhanced training method that uses
TopicPrefix for better awareness of facts and sentence completion as the
training objective, which can vastly reduce the factual errors. We release our
code and FactualityPrompts benchmark at:
https://github.com/nayeon7lee/FactualityPrompt.
| [
{
"version": "v1",
"created": "Thu, 9 Jun 2022 17:16:43 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Oct 2022 06:33:06 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 09:11:34 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lee",
"Nayeon",
""
],
[
"Ping",
"Wei",
""
],
[
"Xu",
"Peng",
""
],
[
"Patwary",
"Mostofa",
""
],
[
"Fung",
"Pascale",
""
],
[
"Shoeybi",
"Mohammad",
""
],
[
"Catanzaro",
"Bryan",
""
]
] | TITLE: Factuality Enhanced Language Models for Open-Ended Text Generation
ABSTRACT: Pretrained language models (LMs) are susceptible to generate text with
nonfactual information. In this work, we measure and improve the factual
accuracy of large-scale LMs for open-ended text generation. We design the
FactualityPrompts test set and metrics to measure the factuality of LM
generations. Based on that, we study the factual accuracy of LMs with parameter
sizes ranging from 126M to 530B. Interestingly, we find that larger LMs are
more factual than smaller ones, although a previous study suggests that larger
LMs can be less truthful in terms of misconceptions. In addition, popular
sampling algorithms (e.g., top-p) in open-ended text generation can harm the
factuality due to the ''uniform randomness'' introduced at every sampling step.
We propose the factual-nucleus sampling algorithm that dynamically adapts the
randomness to improve the factuality of generation while maintaining quality.
Furthermore, we analyze the inefficiencies of the standard training method in
learning correct associations between entities from factual text corpus (e.g.,
Wikipedia). We propose a factuality-enhanced training method that uses
TopicPrefix for better awareness of facts and sentence completion as the
training objective, which can vastly reduce the factual errors. We release our
code and FactualityPrompts benchmark at:
https://github.com/nayeon7lee/FactualityPrompt.
|
2206.06487 | Zihui Xue | Zihui Xue, Zhengqi Gao, Sucheng Ren, Hang Zhao | The Modality Focusing Hypothesis: Towards Understanding Crossmodal
Knowledge Distillation | Accepted by ICLR 2023 (top-5%). The first three authors contribute
equally. Project website: https://zihuixue.github.io/MFH/index.html | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Crossmodal knowledge distillation (KD) extends traditional knowledge
distillation to the area of multimodal learning and demonstrates great success
in various applications. To achieve knowledge transfer across modalities, a
pretrained network from one modality is adopted as the teacher to provide
supervision signals to a student network learning from another modality. In
contrast to the empirical success reported in prior works, the working
mechanism of crossmodal KD remains a mystery. In this paper, we present a
thorough understanding of crossmodal KD. We begin with two case studies and
demonstrate that KD is not a universal cure in crossmodal knowledge transfer.
We then present the modality Venn diagram to understand modality relationships
and the modality focusing hypothesis revealing the decisive factor in the
efficacy of crossmodal KD. Experimental results on 6 multimodal datasets help
justify our hypothesis, diagnose failure cases, and point directions to improve
crossmodal knowledge transfer in the future.
| [
{
"version": "v1",
"created": "Mon, 13 Jun 2022 21:34:21 GMT"
},
{
"version": "v2",
"created": "Fri, 30 Sep 2022 16:43:43 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 01:12:05 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Xue",
"Zihui",
""
],
[
"Gao",
"Zhengqi",
""
],
[
"Ren",
"Sucheng",
""
],
[
"Zhao",
"Hang",
""
]
] | TITLE: The Modality Focusing Hypothesis: Towards Understanding Crossmodal
Knowledge Distillation
ABSTRACT: Crossmodal knowledge distillation (KD) extends traditional knowledge
distillation to the area of multimodal learning and demonstrates great success
in various applications. To achieve knowledge transfer across modalities, a
pretrained network from one modality is adopted as the teacher to provide
supervision signals to a student network learning from another modality. In
contrast to the empirical success reported in prior works, the working
mechanism of crossmodal KD remains a mystery. In this paper, we present a
thorough understanding of crossmodal KD. We begin with two case studies and
demonstrate that KD is not a universal cure in crossmodal knowledge transfer.
We then present the modality Venn diagram to understand modality relationships
and the modality focusing hypothesis revealing the decisive factor in the
efficacy of crossmodal KD. Experimental results on 6 multimodal datasets help
justify our hypothesis, diagnose failure cases, and point directions to improve
crossmodal knowledge transfer in the future.
|
2206.07766 | Yongqiang Chen | Yongqiang Chen, Kaiwen Zhou, Yatao Bian, Binghui Xie, Bingzhe Wu,
Yonggang Zhang, Kaili Ma, Han Yang, Peilin Zhao, Bo Han, James Cheng | Pareto Invariant Risk Minimization: Towards Mitigating the Optimization
Dilemma in Out-of-Distribution Generalization | ICLR 2023, 50 pages, 58 figures | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Recently, there has been a growing surge of interest in enabling machine
learning systems to generalize well to Out-of-Distribution (OOD) data. Most
efforts are devoted to advancing optimization objectives that regularize models
to capture the underlying invariance; however, there often are compromises in
the optimization process of these OOD objectives: i) Many OOD objectives have
to be relaxed as penalty terms of Empirical Risk Minimization (ERM) for the
ease of optimization, while the relaxed forms can weaken the robustness of the
original objective; ii) The penalty terms also require careful tuning of the
penalty weights due to the intrinsic conflicts between ERM and OOD objectives.
Consequently, these compromises could easily lead to suboptimal performance of
either the ERM or OOD objective. To address these issues, we introduce a
multi-objective optimization (MOO) perspective to understand the OOD
optimization process, and propose a new optimization scheme called PAreto
Invariant Risk Minimization (PAIR). PAIR improves the robustness of OOD
objectives by cooperatively optimizing with other OOD objectives, thereby
bridging the gaps caused by the relaxations. Then PAIR approaches a Pareto
optimal solution that trades off the ERM and OOD objectives properly. Extensive
experiments on challenging benchmarks, WILDS, show that PAIR alleviates the
compromises and yields top OOD performances.
| [
{
"version": "v1",
"created": "Wed, 15 Jun 2022 19:04:02 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 09:43:58 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chen",
"Yongqiang",
""
],
[
"Zhou",
"Kaiwen",
""
],
[
"Bian",
"Yatao",
""
],
[
"Xie",
"Binghui",
""
],
[
"Wu",
"Bingzhe",
""
],
[
"Zhang",
"Yonggang",
""
],
[
"Ma",
"Kaili",
""
],
[
"Yang",
"Han",
""
],
[
"Zhao",
"Peilin",
""
],
[
"Han",
"Bo",
""
],
[
"Cheng",
"James",
""
]
] | TITLE: Pareto Invariant Risk Minimization: Towards Mitigating the Optimization
Dilemma in Out-of-Distribution Generalization
ABSTRACT: Recently, there has been a growing surge of interest in enabling machine
learning systems to generalize well to Out-of-Distribution (OOD) data. Most
efforts are devoted to advancing optimization objectives that regularize models
to capture the underlying invariance; however, there often are compromises in
the optimization process of these OOD objectives: i) Many OOD objectives have
to be relaxed as penalty terms of Empirical Risk Minimization (ERM) for the
ease of optimization, while the relaxed forms can weaken the robustness of the
original objective; ii) The penalty terms also require careful tuning of the
penalty weights due to the intrinsic conflicts between ERM and OOD objectives.
Consequently, these compromises could easily lead to suboptimal performance of
either the ERM or OOD objective. To address these issues, we introduce a
multi-objective optimization (MOO) perspective to understand the OOD
optimization process, and propose a new optimization scheme called PAreto
Invariant Risk Minimization (PAIR). PAIR improves the robustness of OOD
objectives by cooperatively optimizing with other OOD objectives, thereby
bridging the gaps caused by the relaxations. Then PAIR approaches a Pareto
optimal solution that trades off the ERM and OOD objectives properly. Extensive
experiments on challenging benchmarks, WILDS, show that PAIR alleviates the
compromises and yields top OOD performances.
|
2206.09034 | Leo Feng | Leo Feng, Mohamed Osama Ahmed, Hossein Hajimirsadeghi, Amir Abdi | Towards Better Selective Classification | null | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We tackle the problem of Selective Classification where the objective is to
achieve the best performance on a predetermined ratio (coverage) of the
dataset. Recent state-of-the-art selective methods come with architectural
changes either via introducing a separate selection head or an extra abstention
logit. In this paper, we challenge the aforementioned methods. The results
suggest that the superior performance of state-of-the-art methods is owed to
training a more generalizable classifier rather than their proposed selection
mechanisms. We argue that the best performing selection mechanism should
instead be rooted in the classifier itself. Our proposed selection strategy
uses the classification scores and achieves better results by a significant
margin, consistently, across all coverages and all datasets, without any added
compute cost. Furthermore, inspired by semi-supervised learning, we propose an
entropy-based regularizer that improves the performance of selective
classification methods. Our proposed selection mechanism with the proposed
entropy-based regularizer achieves new state-of-the-art results.
| [
{
"version": "v1",
"created": "Fri, 17 Jun 2022 22:23:11 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Nov 2022 19:42:20 GMT"
},
{
"version": "v3",
"created": "Fri, 24 Feb 2023 23:41:36 GMT"
},
{
"version": "v4",
"created": "Wed, 1 Mar 2023 20:36:43 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Feng",
"Leo",
""
],
[
"Ahmed",
"Mohamed Osama",
""
],
[
"Hajimirsadeghi",
"Hossein",
""
],
[
"Abdi",
"Amir",
""
]
] | TITLE: Towards Better Selective Classification
ABSTRACT: We tackle the problem of Selective Classification where the objective is to
achieve the best performance on a predetermined ratio (coverage) of the
dataset. Recent state-of-the-art selective methods come with architectural
changes either via introducing a separate selection head or an extra abstention
logit. In this paper, we challenge the aforementioned methods. The results
suggest that the superior performance of state-of-the-art methods is owed to
training a more generalizable classifier rather than their proposed selection
mechanisms. We argue that the best performing selection mechanism should
instead be rooted in the classifier itself. Our proposed selection strategy
uses the classification scores and achieves better results by a significant
margin, consistently, across all coverages and all datasets, without any added
compute cost. Furthermore, inspired by semi-supervised learning, we propose an
entropy-based regularizer that improves the performance of selective
classification methods. Our proposed selection mechanism with the proposed
entropy-based regularizer achieves new state-of-the-art results.
|
2206.09670 | Guiliang Liu | Guiliang Liu, Yudong Luo, Ashish Gaurav, Kasra Rezaee and Pascal
Poupart | Benchmarking Constraint Inference in Inverse Reinforcement Learning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | When deploying Reinforcement Learning (RL) agents into a physical system, we
must ensure that these agents are well aware of the underlying constraints. In
many real-world problems, however, the constraints are often hard to specify
mathematically and unknown to the RL agents. To tackle these issues, Inverse
Constrained Reinforcement Learning (ICRL) empirically estimates constraints
from expert demonstrations. As an emerging research topic, ICRL does not have
common benchmarks, and previous works tested algorithms under hand-crafted
environments with manually-generated expert demonstrations. In this paper, we
construct an ICRL benchmark in the context of RL application domains, including
robot control, and autonomous driving. For each environment, we design relevant
constraints and train expert agents to generate demonstration data. Besides,
unlike existing baselines that learn a deterministic constraint, we propose a
variational ICRL method to model a posterior distribution of candidate
constraints. We conduct extensive experiments on these algorithms under our
benchmark and show how they can facilitate studying important research
challenges for ICRL. The benchmark, including the instructions for reproducing
ICRL algorithms, is available at
https://github.com/Guiliang/ICRL-benchmarks-public.
| [
{
"version": "v1",
"created": "Mon, 20 Jun 2022 09:22:20 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Feb 2023 06:45:07 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 07:20:08 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Liu",
"Guiliang",
""
],
[
"Luo",
"Yudong",
""
],
[
"Gaurav",
"Ashish",
""
],
[
"Rezaee",
"Kasra",
""
],
[
"Poupart",
"Pascal",
""
]
] | TITLE: Benchmarking Constraint Inference in Inverse Reinforcement Learning
ABSTRACT: When deploying Reinforcement Learning (RL) agents into a physical system, we
must ensure that these agents are well aware of the underlying constraints. In
many real-world problems, however, the constraints are often hard to specify
mathematically and unknown to the RL agents. To tackle these issues, Inverse
Constrained Reinforcement Learning (ICRL) empirically estimates constraints
from expert demonstrations. As an emerging research topic, ICRL does not have
common benchmarks, and previous works tested algorithms under hand-crafted
environments with manually-generated expert demonstrations. In this paper, we
construct an ICRL benchmark in the context of RL application domains, including
robot control, and autonomous driving. For each environment, we design relevant
constraints and train expert agents to generate demonstration data. Besides,
unlike existing baselines that learn a deterministic constraint, we propose a
variational ICRL method to model a posterior distribution of candidate
constraints. We conduct extensive experiments on these algorithms under our
benchmark and show how they can facilitate studying important research
challenges for ICRL. The benchmark, including the instructions for reproducing
ICRL algorithms, is available at
https://github.com/Guiliang/ICRL-benchmarks-public.
|
2206.12694 | Yiqing Shen | Yiqing Shen, Yulin Luo, Dinggang Shen, Jing Ke | RandStainNA: Learning Stain-Agnostic Features from Histology Slides by
Bridging Stain Augmentation and Normalization | 12 pages | null | 10.1007/978-3-031-16434-7_21 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Stain variations often decrease the generalization ability of deep learning
based approaches in digital histopathology analysis. Two separate proposals,
namely stain normalization (SN) and stain augmentation (SA), have been
spotlighted to reduce the generalization error, where the former alleviates the
stain shift across different medical centers using template image and the
latter enriches the accessible stain styles by the simulation of more stain
variations. However, their applications are bounded by the selection of
template images and the construction of unrealistic styles. To address the
problems, we unify SN and SA with a novel RandStainNA scheme, which constrains
variable stain styles in a practicable range to train a stain agnostic deep
learning model. The RandStainNA is applicable to stain normalization in a
collection of color spaces i.e. HED, HSV, LAB. Additionally, we propose a
random color space selection scheme to gain extra performance improvement. We
evaluate our method by two diagnostic tasks i.e. tissue subtype classification
and nuclei segmentation, with various network backbones. The performance
superiority over both SA and SN yields that the proposed RandStainNA can
consistently improve the generalization ability, that our models can cope with
more incoming clinical datasets with unpredicted stain styles. The codes is
available at https://github.com/yiqings/RandStainNA.
| [
{
"version": "v1",
"created": "Sat, 25 Jun 2022 16:43:59 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shen",
"Yiqing",
""
],
[
"Luo",
"Yulin",
""
],
[
"Shen",
"Dinggang",
""
],
[
"Ke",
"Jing",
""
]
] | TITLE: RandStainNA: Learning Stain-Agnostic Features from Histology Slides by
Bridging Stain Augmentation and Normalization
ABSTRACT: Stain variations often decrease the generalization ability of deep learning
based approaches in digital histopathology analysis. Two separate proposals,
namely stain normalization (SN) and stain augmentation (SA), have been
spotlighted to reduce the generalization error, where the former alleviates the
stain shift across different medical centers using template image and the
latter enriches the accessible stain styles by the simulation of more stain
variations. However, their applications are bounded by the selection of
template images and the construction of unrealistic styles. To address the
problems, we unify SN and SA with a novel RandStainNA scheme, which constrains
variable stain styles in a practicable range to train a stain agnostic deep
learning model. The RandStainNA is applicable to stain normalization in a
collection of color spaces i.e. HED, HSV, LAB. Additionally, we propose a
random color space selection scheme to gain extra performance improvement. We
evaluate our method by two diagnostic tasks i.e. tissue subtype classification
and nuclei segmentation, with various network backbones. The performance
superiority over both SA and SN yields that the proposed RandStainNA can
consistently improve the generalization ability, that our models can cope with
more incoming clinical datasets with unpredicted stain styles. The codes is
available at https://github.com/yiqings/RandStainNA.
|
2207.00193 | Mingkun Yang | Mingkun Yang, Minghui Liao, Pu Lu, Jing Wang, Shenggao Zhu, Hualin
Luo, Qi Tian, Xiang Bai | Reading and Writing: Discriminative and Generative Modeling for
Self-Supervised Text Recognition | Accepted by ACM MM 2022. The code is available at
https://github.com/ayumiymk/DiG | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing text recognition methods usually need large-scale training data.
Most of them rely on synthetic training data due to the lack of annotated real
images. However, there is a domain gap between the synthetic data and real
data, which limits the performance of the text recognition models. Recent
self-supervised text recognition methods attempted to utilize unlabeled real
images by introducing contrastive learning, which mainly learns the
discrimination of the text images. Inspired by the observation that humans
learn to recognize the texts through both reading and writing, we propose to
learn discrimination and generation by integrating contrastive learning and
masked image modeling in our self-supervised method. The contrastive learning
branch is adopted to learn the discrimination of text images, which imitates
the reading behavior of humans. Meanwhile, masked image modeling is firstly
introduced for text recognition to learn the context generation of the text
images, which is similar to the writing behavior. The experimental results show
that our method outperforms previous self-supervised text recognition methods
by 10.2%-20.2% on irregular scene text recognition datasets. Moreover, our
proposed text recognizer exceeds previous state-of-the-art text recognition
methods by averagely 5.3% on 11 benchmarks, with similar model size. We also
demonstrate that our pre-trained model can be easily applied to other
text-related tasks with obvious performance gain. The code is available at
https://github.com/ayumiymk/DiG.
| [
{
"version": "v1",
"created": "Fri, 1 Jul 2022 03:50:26 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 07:39:00 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yang",
"Mingkun",
""
],
[
"Liao",
"Minghui",
""
],
[
"Lu",
"Pu",
""
],
[
"Wang",
"Jing",
""
],
[
"Zhu",
"Shenggao",
""
],
[
"Luo",
"Hualin",
""
],
[
"Tian",
"Qi",
""
],
[
"Bai",
"Xiang",
""
]
] | TITLE: Reading and Writing: Discriminative and Generative Modeling for
Self-Supervised Text Recognition
ABSTRACT: Existing text recognition methods usually need large-scale training data.
Most of them rely on synthetic training data due to the lack of annotated real
images. However, there is a domain gap between the synthetic data and real
data, which limits the performance of the text recognition models. Recent
self-supervised text recognition methods attempted to utilize unlabeled real
images by introducing contrastive learning, which mainly learns the
discrimination of the text images. Inspired by the observation that humans
learn to recognize the texts through both reading and writing, we propose to
learn discrimination and generation by integrating contrastive learning and
masked image modeling in our self-supervised method. The contrastive learning
branch is adopted to learn the discrimination of text images, which imitates
the reading behavior of humans. Meanwhile, masked image modeling is firstly
introduced for text recognition to learn the context generation of the text
images, which is similar to the writing behavior. The experimental results show
that our method outperforms previous self-supervised text recognition methods
by 10.2%-20.2% on irregular scene text recognition datasets. Moreover, our
proposed text recognizer exceeds previous state-of-the-art text recognition
methods by averagely 5.3% on 11 benchmarks, with similar model size. We also
demonstrate that our pre-trained model can be easily applied to other
text-related tasks with obvious performance gain. The code is available at
https://github.com/ayumiymk/DiG.
|
2207.00215 | Zhuoxiao Li | Zhuoxiao Li, Haiyang Jiang, Yinqiang Zheng | Polarized Color Image Denoising using Pocoformer | New version is accpeted by CVPR 2023 and great modifications are
taken | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Polarized color photography provides both visual textures and object
surficial information in one single snapshot. However, the use of the
directional polarizing filter array causes extremely lower photon count and SNR
compared to conventional color imaging. Thus, the feature essentially leads to
unpleasant noisy images and destroys polarization analysis performance. It is a
challenge for traditional image processing pipelines owing to the fact that the
physical constraints exerted implicitly in the channels are excessively
complicated. To address this issue, we propose a learning-based approach to
simultaneously restore clean signals and precise polarization information. A
real-world polarized color image dataset of paired raw short-exposed noisy and
long-exposed reference images are captured to support the learning-based
pipeline. Moreover, we embrace the development of vision Transformer and
propose a hybrid transformer model for the Polarized Color image denoising,
namely PoCoformer, for a better restoration performance. Abundant experiments
demonstrate the effectiveness of proposed method and key factors that affect
results are analyzed.
| [
{
"version": "v1",
"created": "Fri, 1 Jul 2022 05:52:14 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 03:07:47 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Li",
"Zhuoxiao",
""
],
[
"Jiang",
"Haiyang",
""
],
[
"Zheng",
"Yinqiang",
""
]
] | TITLE: Polarized Color Image Denoising using Pocoformer
ABSTRACT: Polarized color photography provides both visual textures and object
surficial information in one single snapshot. However, the use of the
directional polarizing filter array causes extremely lower photon count and SNR
compared to conventional color imaging. Thus, the feature essentially leads to
unpleasant noisy images and destroys polarization analysis performance. It is a
challenge for traditional image processing pipelines owing to the fact that the
physical constraints exerted implicitly in the channels are excessively
complicated. To address this issue, we propose a learning-based approach to
simultaneously restore clean signals and precise polarization information. A
real-world polarized color image dataset of paired raw short-exposed noisy and
long-exposed reference images are captured to support the learning-based
pipeline. Moreover, we embrace the development of vision Transformer and
propose a hybrid transformer model for the Polarized Color image denoising,
namely PoCoformer, for a better restoration performance. Abundant experiments
demonstrate the effectiveness of proposed method and key factors that affect
results are analyzed.
|
2207.04159 | Matthijs Jansen | Matthijs Jansen, Auday Al-Dulaimy, Alessandro V. Papadopoulos, Animesh
Trivedi, Alexandru Iosup | The SPEC-RG Reference Architecture for the Compute Continuum | 14 pages, SPEC-RG technical report | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As the next generation of diverse workloads like autonomous driving and
augmented/virtual reality evolves, computation is shifting from cloud-based
services to the edge, leading to the emergence of a cloud-edge compute
continuum. This continuum promises a wide spectrum of deployment opportunities
for workloads that can leverage the strengths of cloud (scalable
infrastructure, high reliability) and edge (energy efficient, low latencies).
Despite its promises, the continuum has only been studied in silos of various
computing models, thus lacking strong end-to-end theoretical and engineering
foundations for computing and resource management across the continuum.
Consequently, developers resort to ad hoc approaches to reason about
performance and resource utilization of workloads in the continuum. In this
work, we conduct a first-of-its-kind systematic study of various computing
models, identify salient properties, and make a case to unify them under a
compute continuum reference architecture. This architecture provides an
end-to-end analysis framework for developers to reason about resource
management, workload distribution, and performance analysis. We demonstrate the
utility of the reference architecture by analyzing two popular continuum
workloads, deep learning and industrial IoT. We have developed an accompanying
deployment and benchmarking framework and first-order analytical model for
quantitative reasoning of continuum workloads. The framework is open-sourced
and available at https://github.com/atlarge-research/continuum.
| [
{
"version": "v1",
"created": "Sat, 9 Jul 2022 00:05:16 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Dec 2022 20:12:03 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 13:26:26 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Jansen",
"Matthijs",
""
],
[
"Al-Dulaimy",
"Auday",
""
],
[
"Papadopoulos",
"Alessandro V.",
""
],
[
"Trivedi",
"Animesh",
""
],
[
"Iosup",
"Alexandru",
""
]
] | TITLE: The SPEC-RG Reference Architecture for the Compute Continuum
ABSTRACT: As the next generation of diverse workloads like autonomous driving and
augmented/virtual reality evolves, computation is shifting from cloud-based
services to the edge, leading to the emergence of a cloud-edge compute
continuum. This continuum promises a wide spectrum of deployment opportunities
for workloads that can leverage the strengths of cloud (scalable
infrastructure, high reliability) and edge (energy efficient, low latencies).
Despite its promises, the continuum has only been studied in silos of various
computing models, thus lacking strong end-to-end theoretical and engineering
foundations for computing and resource management across the continuum.
Consequently, developers resort to ad hoc approaches to reason about
performance and resource utilization of workloads in the continuum. In this
work, we conduct a first-of-its-kind systematic study of various computing
models, identify salient properties, and make a case to unify them under a
compute continuum reference architecture. This architecture provides an
end-to-end analysis framework for developers to reason about resource
management, workload distribution, and performance analysis. We demonstrate the
utility of the reference architecture by analyzing two popular continuum
workloads, deep learning and industrial IoT. We have developed an accompanying
deployment and benchmarking framework and first-order analytical model for
quantitative reasoning of continuum workloads. The framework is open-sourced
and available at https://github.com/atlarge-research/continuum.
|
2207.07027 | Farah Shamout | Nasir Hayat, Krzysztof J. Geras, Farah E. Shamout | MedFuse: Multi-modal fusion with clinical time-series data and chest
X-ray images | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multi-modal fusion approaches aim to integrate information from different
data sources. Unlike natural datasets, such as in audio-visual applications,
where samples consist of "paired" modalities, data in healthcare is often
collected asynchronously. Hence, requiring the presence of all modalities for a
given sample is not realistic for clinical tasks and significantly limits the
size of the dataset during training. In this paper, we propose MedFuse, a
conceptually simple yet promising LSTM-based fusion module that can accommodate
uni-modal as well as multi-modal input. We evaluate the fusion method and
introduce new benchmark results for in-hospital mortality prediction and
phenotype classification, using clinical time-series data in the MIMIC-IV
dataset and corresponding chest X-ray images in MIMIC-CXR. Compared to more
complex multi-modal fusion strategies, MedFuse provides a performance
improvement by a large margin on the fully paired test set. It also remains
robust across the partially paired test set containing samples with missing
chest X-ray images. We release our code for reproducibility and to enable the
evaluation of competing models in the future.
| [
{
"version": "v1",
"created": "Thu, 14 Jul 2022 15:59:03 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 14:49:06 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Hayat",
"Nasir",
""
],
[
"Geras",
"Krzysztof J.",
""
],
[
"Shamout",
"Farah E.",
""
]
] | TITLE: MedFuse: Multi-modal fusion with clinical time-series data and chest
X-ray images
ABSTRACT: Multi-modal fusion approaches aim to integrate information from different
data sources. Unlike natural datasets, such as in audio-visual applications,
where samples consist of "paired" modalities, data in healthcare is often
collected asynchronously. Hence, requiring the presence of all modalities for a
given sample is not realistic for clinical tasks and significantly limits the
size of the dataset during training. In this paper, we propose MedFuse, a
conceptually simple yet promising LSTM-based fusion module that can accommodate
uni-modal as well as multi-modal input. We evaluate the fusion method and
introduce new benchmark results for in-hospital mortality prediction and
phenotype classification, using clinical time-series data in the MIMIC-IV
dataset and corresponding chest X-ray images in MIMIC-CXR. Compared to more
complex multi-modal fusion strategies, MedFuse provides a performance
improvement by a large margin on the fully paired test set. It also remains
robust across the partially paired test set containing samples with missing
chest X-ray images. We release our code for reproducibility and to enable the
evaluation of competing models in the future.
|
2208.01174 | Peter Jansen | Peter A. Jansen, Marc-Alexandre C\^ot\'e | TextWorldExpress: Simulating Text Games at One Million Steps Per Second | Accepted to EACL 2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Text-based games offer a challenging test bed to evaluate virtual agents at
language understanding, multi-step problem-solving, and common-sense reasoning.
However, speed is a major limitation of current text-based games, capping at
300 steps per second, mainly due to the use of legacy tooling. In this work we
present TextWorldExpress, a high-performance simulator that includes
implementations of three common text game benchmarks that increases simulation
throughput by approximately three orders of magnitude, reaching over one
million steps per second on common desktop hardware. This significantly reduces
experiment runtime, enabling billion-step-scale experiments in about one day.
| [
{
"version": "v1",
"created": "Mon, 1 Aug 2022 23:43:48 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 06:11:49 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Jansen",
"Peter A.",
""
],
[
"Côté",
"Marc-Alexandre",
""
]
] | TITLE: TextWorldExpress: Simulating Text Games at One Million Steps Per Second
ABSTRACT: Text-based games offer a challenging test bed to evaluate virtual agents at
language understanding, multi-step problem-solving, and common-sense reasoning.
However, speed is a major limitation of current text-based games, capping at
300 steps per second, mainly due to the use of legacy tooling. In this work we
present TextWorldExpress, a high-performance simulator that includes
implementations of three common text game benchmarks that increases simulation
throughput by approximately three orders of magnitude, reaching over one
million steps per second on common desktop hardware. This significantly reduces
experiment runtime, enabling billion-step-scale experiments in about one day.
|
2208.01214 | Jun Xue | Jun Xue, Cunhang Fan, Zhao Lv, Jianhua Tao, Jiangyan Yi, Chengshi
Zheng, Zhengqi Wen, Minmin Yuan, Shegang Shao | Audio Deepfake Detection Based on a Combination of F0 Information and
Real Plus Imaginary Spectrogram Features | null | null | 10.1145/3552466.3556526 | null | cs.SD cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, pioneer research works have proposed a large number of acoustic
features (log power spectrogram, linear frequency cepstral coefficients,
constant Q cepstral coefficients, etc.) for audio deepfake detection, obtaining
good performance, and showing that different subbands have different
contributions to audio deepfake detection. However, this lacks an explanation
of the specific information in the subband, and these features also lose
information such as phase. Inspired by the mechanism of synthetic speech, the
fundamental frequency (F0) information is used to improve the quality of
synthetic speech, while the F0 of synthetic speech is still too average, which
differs significantly from that of real speech. It is expected that F0 can be
used as important information to discriminate between bonafide and fake speech,
while this information cannot be used directly due to the irregular
distribution of F0. Insteadly, the frequency band containing most of F0 is
selected as the input feature. Meanwhile, to make full use of the phase and
full-band information, we also propose to use real and imaginary spectrogram
features as complementary input features and model the disjoint subbands
separately. Finally, the results of F0, real and imaginary spectrogram features
are fused. Experimental results on the ASVspoof 2019 LA dataset show that our
proposed system is very effective for the audio deepfake detection task,
achieving an equivalent error rate (EER) of 0.43%, which surpasses almost all
systems.
| [
{
"version": "v1",
"created": "Tue, 2 Aug 2022 02:46:16 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Xue",
"Jun",
""
],
[
"Fan",
"Cunhang",
""
],
[
"Lv",
"Zhao",
""
],
[
"Tao",
"Jianhua",
""
],
[
"Yi",
"Jiangyan",
""
],
[
"Zheng",
"Chengshi",
""
],
[
"Wen",
"Zhengqi",
""
],
[
"Yuan",
"Minmin",
""
],
[
"Shao",
"Shegang",
""
]
] | TITLE: Audio Deepfake Detection Based on a Combination of F0 Information and
Real Plus Imaginary Spectrogram Features
ABSTRACT: Recently, pioneer research works have proposed a large number of acoustic
features (log power spectrogram, linear frequency cepstral coefficients,
constant Q cepstral coefficients, etc.) for audio deepfake detection, obtaining
good performance, and showing that different subbands have different
contributions to audio deepfake detection. However, this lacks an explanation
of the specific information in the subband, and these features also lose
information such as phase. Inspired by the mechanism of synthetic speech, the
fundamental frequency (F0) information is used to improve the quality of
synthetic speech, while the F0 of synthetic speech is still too average, which
differs significantly from that of real speech. It is expected that F0 can be
used as important information to discriminate between bonafide and fake speech,
while this information cannot be used directly due to the irregular
distribution of F0. Insteadly, the frequency band containing most of F0 is
selected as the input feature. Meanwhile, to make full use of the phase and
full-band information, we also propose to use real and imaginary spectrogram
features as complementary input features and model the disjoint subbands
separately. Finally, the results of F0, real and imaginary spectrogram features
are fused. Experimental results on the ASVspoof 2019 LA dataset show that our
proposed system is very effective for the audio deepfake detection task,
achieving an equivalent error rate (EER) of 0.43%, which surpasses almost all
systems.
|
2208.03028 | Wei Dai | Wei Dai, Ziyao Zhang, Lixia Tian, Shengyuan Yu, Shuhui Wang, Zhao
Dong, and Hairong Zheng | Multimodal Brain Disease Classification with Functional Interaction
Learning from Single fMRI Volume | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | In neuroimaging analysis, fMRI can well assess the function changes for brain
diseases with no obvious structural lesions. To date, most deep-learning-based
fMRI studies have employed functional connectivity (FC) as the basic feature
for disease classification. However, FC is calculated on time series of
predefined regions of interest and neglects detailed information contained in
each voxel. Another drawback of using FC is the limited sample size for the
training of deep models. The low representation ability of FC leads to poor
performance in clinical practice, especially when dealing with multimodal
medical data involving multiple types of visual signals and textual records for
brain diseases. To overcome this bottleneck problem in the fMRI feature
modality, we propose BrainFormer, an end-to-end functional interaction learning
method for brain disease classification with single fMRI volume. Unlike
traditional deep learning methods that construct convolution and transformers
on FC, BrainFormer learns the functional interaction from fMRI signals, by
modeling the local cues within each voxel with 3D convolutions and capturing
the global correlations among distant regions with specially designed global
attention mechanisms from shallow layers to deep layers. Meanwhile, BrainFormer
can deal with multimodal medical data including fMRI volume, structural MRI, FC
features and phenotypic data to achieve more comprehensive brain disease
diagnosis. We evaluate BrainFormer on five independent multi-site datasets on
autism, Alzheimer's disease, depression, attention deficit hyperactivity
disorder and headache disorders. The results demonstrate its effectiveness and
generalizability for multiple brain diseases diagnosis with multimodal
features. BrainFormer may promote precision of neuroimaging-based diagnosis in
clinical practice and motivate future studies on fMRI analysis.
| [
{
"version": "v1",
"created": "Fri, 5 Aug 2022 07:54:10 GMT"
},
{
"version": "v2",
"created": "Wed, 10 Aug 2022 05:41:13 GMT"
},
{
"version": "v3",
"created": "Wed, 1 Mar 2023 06:54:58 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Dai",
"Wei",
""
],
[
"Zhang",
"Ziyao",
""
],
[
"Tian",
"Lixia",
""
],
[
"Yu",
"Shengyuan",
""
],
[
"Wang",
"Shuhui",
""
],
[
"Dong",
"Zhao",
""
],
[
"Zheng",
"Hairong",
""
]
] | TITLE: Multimodal Brain Disease Classification with Functional Interaction
Learning from Single fMRI Volume
ABSTRACT: In neuroimaging analysis, fMRI can well assess the function changes for brain
diseases with no obvious structural lesions. To date, most deep-learning-based
fMRI studies have employed functional connectivity (FC) as the basic feature
for disease classification. However, FC is calculated on time series of
predefined regions of interest and neglects detailed information contained in
each voxel. Another drawback of using FC is the limited sample size for the
training of deep models. The low representation ability of FC leads to poor
performance in clinical practice, especially when dealing with multimodal
medical data involving multiple types of visual signals and textual records for
brain diseases. To overcome this bottleneck problem in the fMRI feature
modality, we propose BrainFormer, an end-to-end functional interaction learning
method for brain disease classification with single fMRI volume. Unlike
traditional deep learning methods that construct convolution and transformers
on FC, BrainFormer learns the functional interaction from fMRI signals, by
modeling the local cues within each voxel with 3D convolutions and capturing
the global correlations among distant regions with specially designed global
attention mechanisms from shallow layers to deep layers. Meanwhile, BrainFormer
can deal with multimodal medical data including fMRI volume, structural MRI, FC
features and phenotypic data to achieve more comprehensive brain disease
diagnosis. We evaluate BrainFormer on five independent multi-site datasets on
autism, Alzheimer's disease, depression, attention deficit hyperactivity
disorder and headache disorders. The results demonstrate its effectiveness and
generalizability for multiple brain diseases diagnosis with multimodal
features. BrainFormer may promote precision of neuroimaging-based diagnosis in
clinical practice and motivate future studies on fMRI analysis.
|
2208.13224 | Florian Putz | Thomas Weissmann, Yixing Huang, Stefan Fischer, Johannes Roesch, Sina
Mansoorian, Horacio Ayala Gaona, Antoniu-Oreste Gostian, Markus Hecht,
Sebastian Lettmaier, Lisa Deloch, Benjamin Frey, Udo S. Gaipl, Luitpold V.
Distel, Andreas Maier, Heinrich Iro, Sabine Semrau, Christoph Bert, Rainer
Fietkau and Florian Putz | Deep learning for automatic head and neck lymph node level delineation
provides expert-level accuracy | 14 pages, 6 figures, published in Frontiers in Oncology | Front. Oncol. 13:1115258 | 10.3389/fonc.2023.1115258 | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Background: Deep learning (DL)-based head and neck lymph node level (HN_LNL)
autodelineation is of high relevance to radiotherapy research and clinical
treatment planning but still underinvestigated in academic literature. Methods:
An expert-delineated cohort of 35 planning CTs was used for training of an
nnU-net 3D-fullres/2D-ensemble model for autosegmentation of 20 different
HN_LNL. A second cohort acquired at the same institution later in time served
as the test set (n=20). In a completely blinded evaluation, 3 clinical experts
rated the quality of DL autosegmentations in a head-to-head comparison with
expert-created contours. For a subgroup of 10 cases, intraobserver variability
was compared to the average DL autosegmentation accuracy on the original and
recontoured set of expert segmentations. A postprocessing step to adjust
craniocaudal boundaries of level autosegmentations to the CT slice plane was
introduced and the effect on geometric accuracy and expert rating was
investigated. Results: Blinded expert ratings for DL segmentations and
expert-created contours were not significantly different. DL segmentations with
slice plane adjustment were rated numerically higher (mean, 81.0 vs.
79.6,p=0.185) and DL segmentations without slice plane adjustment were rated
numerically lower (77.2 vs. 79.6,p=0.167) than manually drawn contours. DL
segmentations with CT slice plane adjustment were rated significantly better
than DL contours without slice plane adjustment (81.0 vs. 77.2,p=0.004).
Geometric accuracy of DL segmentations was not different from intraobserver
variability (mean, 0.76 vs. 0.77, p=0.307). Conclusions: We show that a nnU-net
3D-fullres/2D-ensemble model can be used for highly accurate autodelineation of
HN_LNL using only a limited training dataset that is ideally suited for
large-scale standardized autodelineation of HN_LNL in the research setting.
| [
{
"version": "v1",
"created": "Sun, 28 Aug 2022 13:58:54 GMT"
},
{
"version": "v2",
"created": "Wed, 1 Mar 2023 19:54:06 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Weissmann",
"Thomas",
""
],
[
"Huang",
"Yixing",
""
],
[
"Fischer",
"Stefan",
""
],
[
"Roesch",
"Johannes",
""
],
[
"Mansoorian",
"Sina",
""
],
[
"Gaona",
"Horacio Ayala",
""
],
[
"Gostian",
"Antoniu-Oreste",
""
],
[
"Hecht",
"Markus",
""
],
[
"Lettmaier",
"Sebastian",
""
],
[
"Deloch",
"Lisa",
""
],
[
"Frey",
"Benjamin",
""
],
[
"Gaipl",
"Udo S.",
""
],
[
"Distel",
"Luitpold V.",
""
],
[
"Maier",
"Andreas",
""
],
[
"Iro",
"Heinrich",
""
],
[
"Semrau",
"Sabine",
""
],
[
"Bert",
"Christoph",
""
],
[
"Fietkau",
"Rainer",
""
],
[
"Putz",
"Florian",
""
]
] | TITLE: Deep learning for automatic head and neck lymph node level delineation
provides expert-level accuracy
ABSTRACT: Background: Deep learning (DL)-based head and neck lymph node level (HN_LNL)
autodelineation is of high relevance to radiotherapy research and clinical
treatment planning but still underinvestigated in academic literature. Methods:
An expert-delineated cohort of 35 planning CTs was used for training of an
nnU-net 3D-fullres/2D-ensemble model for autosegmentation of 20 different
HN_LNL. A second cohort acquired at the same institution later in time served
as the test set (n=20). In a completely blinded evaluation, 3 clinical experts
rated the quality of DL autosegmentations in a head-to-head comparison with
expert-created contours. For a subgroup of 10 cases, intraobserver variability
was compared to the average DL autosegmentation accuracy on the original and
recontoured set of expert segmentations. A postprocessing step to adjust
craniocaudal boundaries of level autosegmentations to the CT slice plane was
introduced and the effect on geometric accuracy and expert rating was
investigated. Results: Blinded expert ratings for DL segmentations and
expert-created contours were not significantly different. DL segmentations with
slice plane adjustment were rated numerically higher (mean, 81.0 vs.
79.6,p=0.185) and DL segmentations without slice plane adjustment were rated
numerically lower (77.2 vs. 79.6,p=0.167) than manually drawn contours. DL
segmentations with CT slice plane adjustment were rated significantly better
than DL contours without slice plane adjustment (81.0 vs. 77.2,p=0.004).
Geometric accuracy of DL segmentations was not different from intraobserver
variability (mean, 0.76 vs. 0.77, p=0.307). Conclusions: We show that a nnU-net
3D-fullres/2D-ensemble model can be used for highly accurate autodelineation of
HN_LNL using only a limited training dataset that is ideally suited for
large-scale standardized autodelineation of HN_LNL in the research setting.
|
2209.04836 | Samuel Ainsworth | Samuel K. Ainsworth, Jonathan Hayase, Siddhartha Srinivasa | Git Re-Basin: Merging Models modulo Permutation Symmetries | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The success of deep learning is due in large part to our ability to solve
certain massive non-convex optimization problems with relative ease. Though
non-convex optimization is NP-hard, simple algorithms -- often variants of
stochastic gradient descent -- exhibit surprising effectiveness in fitting
large neural networks in practice. We argue that neural network loss landscapes
often contain (nearly) a single basin after accounting for all possible
permutation symmetries of hidden units a la Entezari et al. 2021. We introduce
three algorithms to permute the units of one model to bring them into alignment
with a reference model in order to merge the two models in weight space. This
transformation produces a functionally equivalent set of weights that lie in an
approximately convex basin near the reference model. Experimentally, we
demonstrate the single basin phenomenon across a variety of model architectures
and datasets, including the first (to our knowledge) demonstration of
zero-barrier linear mode connectivity between independently trained ResNet
models on CIFAR-10. Additionally, we identify intriguing phenomena relating
model width and training time to mode connectivity. Finally, we discuss
shortcomings of the linear mode connectivity hypothesis, including a
counterexample to the single basin theory.
| [
{
"version": "v1",
"created": "Sun, 11 Sep 2022 10:44:27 GMT"
},
{
"version": "v2",
"created": "Fri, 7 Oct 2022 00:38:25 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Nov 2022 07:14:00 GMT"
},
{
"version": "v4",
"created": "Sun, 4 Dec 2022 23:59:26 GMT"
},
{
"version": "v5",
"created": "Thu, 29 Dec 2022 23:33:35 GMT"
},
{
"version": "v6",
"created": "Wed, 1 Mar 2023 22:17:16 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Ainsworth",
"Samuel K.",
""
],
[
"Hayase",
"Jonathan",
""
],
[
"Srinivasa",
"Siddhartha",
""
]
] | TITLE: Git Re-Basin: Merging Models modulo Permutation Symmetries
ABSTRACT: The success of deep learning is due in large part to our ability to solve
certain massive non-convex optimization problems with relative ease. Though
non-convex optimization is NP-hard, simple algorithms -- often variants of
stochastic gradient descent -- exhibit surprising effectiveness in fitting
large neural networks in practice. We argue that neural network loss landscapes
often contain (nearly) a single basin after accounting for all possible
permutation symmetries of hidden units a la Entezari et al. 2021. We introduce
three algorithms to permute the units of one model to bring them into alignment
with a reference model in order to merge the two models in weight space. This
transformation produces a functionally equivalent set of weights that lie in an
approximately convex basin near the reference model. Experimentally, we
demonstrate the single basin phenomenon across a variety of model architectures
and datasets, including the first (to our knowledge) demonstration of
zero-barrier linear mode connectivity between independently trained ResNet
models on CIFAR-10. Additionally, we identify intriguing phenomena relating
model width and training time to mode connectivity. Finally, we discuss
shortcomings of the linear mode connectivity hypothesis, including a
counterexample to the single basin theory.
|
2209.06430 | Hongwei Xue | Hongwei Xue, Yuchong Sun, Bei Liu, Jianlong Fu, Ruihua Song, Houqiang
Li, Jiebo Luo | CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language
Representation Alignment | Accepted by ICLR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The pre-trained image-text models, like CLIP, have demonstrated the strong
power of vision-language representation learned from a large scale of
web-collected image-text data. In light of the well-learned visual features,
some existing works transfer image representation to video domain and achieve
good results. However, how to utilize image-language pre-trained model (e.g.,
CLIP) for video-language pre-training (post-pretraining) is still under
explored. In this paper, we investigate two questions: 1) what are the factors
hindering post-pretraining CLIP to further improve the performance on
video-language tasks? and 2) how to mitigate the impact of these factors?
Through a series of comparative experiments and analyses, we find that the data
scale and domain gap between language sources have great impacts. Motivated by
these, we propose a Omnisource Cross-modal Learning method equipped with a
Video Proxy mechanism on the basis of CLIP, namely CLIP-ViP. Extensive results
show that our approach improves the performance of CLIP on video-text retrieval
by a large margin. Our model also achieves SOTA results on a variety of
datasets, including MSR-VTT, DiDeMo, LSMDC, and ActivityNet. We will release
our code and pre-trained CLIP-ViP models at
https://github.com/microsoft/XPretrain/tree/main/CLIP-ViP.
| [
{
"version": "v1",
"created": "Wed, 14 Sep 2022 05:47:02 GMT"
},
{
"version": "v2",
"created": "Fri, 23 Sep 2022 03:38:25 GMT"
},
{
"version": "v3",
"created": "Mon, 27 Feb 2023 04:25:24 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Mar 2023 08:24:23 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Xue",
"Hongwei",
""
],
[
"Sun",
"Yuchong",
""
],
[
"Liu",
"Bei",
""
],
[
"Fu",
"Jianlong",
""
],
[
"Song",
"Ruihua",
""
],
[
"Li",
"Houqiang",
""
],
[
"Luo",
"Jiebo",
""
]
] | TITLE: CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language
Representation Alignment
ABSTRACT: The pre-trained image-text models, like CLIP, have demonstrated the strong
power of vision-language representation learned from a large scale of
web-collected image-text data. In light of the well-learned visual features,
some existing works transfer image representation to video domain and achieve
good results. However, how to utilize image-language pre-trained model (e.g.,
CLIP) for video-language pre-training (post-pretraining) is still under
explored. In this paper, we investigate two questions: 1) what are the factors
hindering post-pretraining CLIP to further improve the performance on
video-language tasks? and 2) how to mitigate the impact of these factors?
Through a series of comparative experiments and analyses, we find that the data
scale and domain gap between language sources have great impacts. Motivated by
these, we propose a Omnisource Cross-modal Learning method equipped with a
Video Proxy mechanism on the basis of CLIP, namely CLIP-ViP. Extensive results
show that our approach improves the performance of CLIP on video-text retrieval
by a large margin. Our model also achieves SOTA results on a variety of
datasets, including MSR-VTT, DiDeMo, LSMDC, and ActivityNet. We will release
our code and pre-trained CLIP-ViP models at
https://github.com/microsoft/XPretrain/tree/main/CLIP-ViP.
|
2209.11294 | Benjamin Stoler | Benjamin Stoler, Meghdeep Jana, Soonmin Hwang, Jean Oh | T2FPV: Dataset and Method for Correcting First-Person View Errors in
Pedestrian Trajectory Prediction | null | null | null | null | cs.CV cs.RO | http://creativecommons.org/licenses/by-sa/4.0/ | Predicting pedestrian motion is essential for developing socially-aware
robots that interact in a crowded environment. While the natural visual
perspective for a social interaction setting is an egocentric view, the
majority of existing work in trajectory prediction therein has been
investigated purely in the top-down trajectory space. To support first-person
view trajectory prediction research, we present T2FPV, a method for
constructing high-fidelity first-person view (FPV) datasets given a real-world,
top-down trajectory dataset; we showcase our approach on the ETH/UCY pedestrian
dataset to generate the egocentric visual data of all interacting pedestrians,
creating the T2FPV-ETH dataset. In this setting, FPV-specific errors arise due
to imperfect detection and tracking, occlusions, and field-of-view (FOV)
limitations of the camera. To address these errors, we propose CoFE, a module
that further refines the imputation of missing data in an end-to-end manner
with trajectory forecasting algorithms. Our method reduces the impact of such
FPV errors on downstream prediction performance, decreasing displacement error
by more than 10% on average. To facilitate research engagement, we release our
T2FPV-ETH dataset and software tools.
| [
{
"version": "v1",
"created": "Thu, 22 Sep 2022 20:14:43 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 07:51:07 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Stoler",
"Benjamin",
""
],
[
"Jana",
"Meghdeep",
""
],
[
"Hwang",
"Soonmin",
""
],
[
"Oh",
"Jean",
""
]
] | TITLE: T2FPV: Dataset and Method for Correcting First-Person View Errors in
Pedestrian Trajectory Prediction
ABSTRACT: Predicting pedestrian motion is essential for developing socially-aware
robots that interact in a crowded environment. While the natural visual
perspective for a social interaction setting is an egocentric view, the
majority of existing work in trajectory prediction therein has been
investigated purely in the top-down trajectory space. To support first-person
view trajectory prediction research, we present T2FPV, a method for
constructing high-fidelity first-person view (FPV) datasets given a real-world,
top-down trajectory dataset; we showcase our approach on the ETH/UCY pedestrian
dataset to generate the egocentric visual data of all interacting pedestrians,
creating the T2FPV-ETH dataset. In this setting, FPV-specific errors arise due
to imperfect detection and tracking, occlusions, and field-of-view (FOV)
limitations of the camera. To address these errors, we propose CoFE, a module
that further refines the imputation of missing data in an end-to-end manner
with trajectory forecasting algorithms. Our method reduces the impact of such
FPV errors on downstream prediction performance, decreasing displacement error
by more than 10% on average. To facilitate research engagement, we release our
T2FPV-ETH dataset and software tools.
|
2209.14610 | Pan Lu | Pan Lu, Liang Qiu, Kai-Wei Chang, Ying Nian Wu, Song-Chun Zhu, Tanmay
Rajpurohit, Peter Clark, Ashwin Kalyan | Dynamic Prompt Learning via Policy Gradient for Semi-structured
Mathematical Reasoning | ICLR 2023. 26 pages and 18 figures. The data and code are available
at https://promptpg.github.io | null | null | null | cs.LG cs.AI cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | Mathematical reasoning, a core ability of human intelligence, presents unique
challenges for machines in abstract thinking and logical reasoning. Recent
large pre-trained language models such as GPT-3 have achieved remarkable
progress on mathematical reasoning tasks written in text form, such as math
word problems (MWP). However, it is unknown if the models can handle more
complex problems that involve math reasoning over heterogeneous information,
such as tabular data. To fill the gap, we present Tabular Math Word Problems
(TabMWP), a new dataset containing 38,431 open-domain grade-level problems that
require mathematical reasoning on both textual and tabular data. Each question
in TabMWP is aligned with a tabular context, which is presented as an image,
semi-structured text, and a structured table. There are two types of questions:
free-text and multi-choice, and each problem is annotated with gold solutions
to reveal the multi-step reasoning process. We evaluate different pre-trained
models on TabMWP, including the GPT-3 model in a few-shot setting. As earlier
studies suggest, since few-shot GPT-3 relies on the selection of in-context
examples, its performance is unstable and can degrade to near chance. The
unstable issue is more severe when handling complex problems like TabMWP. To
mitigate this, we further propose a novel approach, PromptPG, which utilizes
policy gradient to learn to select in-context examples from a small amount of
training data and then constructs the corresponding prompt for the test
example. Experimental results show that our method outperforms the best
baseline by 5.31% on the accuracy metric and reduces the prediction variance
significantly compared to random selection, which verifies its effectiveness in
selecting in-context examples.
| [
{
"version": "v1",
"created": "Thu, 29 Sep 2022 08:01:04 GMT"
},
{
"version": "v2",
"created": "Wed, 2 Nov 2022 23:42:14 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 07:41:55 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lu",
"Pan",
""
],
[
"Qiu",
"Liang",
""
],
[
"Chang",
"Kai-Wei",
""
],
[
"Wu",
"Ying Nian",
""
],
[
"Zhu",
"Song-Chun",
""
],
[
"Rajpurohit",
"Tanmay",
""
],
[
"Clark",
"Peter",
""
],
[
"Kalyan",
"Ashwin",
""
]
] | TITLE: Dynamic Prompt Learning via Policy Gradient for Semi-structured
Mathematical Reasoning
ABSTRACT: Mathematical reasoning, a core ability of human intelligence, presents unique
challenges for machines in abstract thinking and logical reasoning. Recent
large pre-trained language models such as GPT-3 have achieved remarkable
progress on mathematical reasoning tasks written in text form, such as math
word problems (MWP). However, it is unknown if the models can handle more
complex problems that involve math reasoning over heterogeneous information,
such as tabular data. To fill the gap, we present Tabular Math Word Problems
(TabMWP), a new dataset containing 38,431 open-domain grade-level problems that
require mathematical reasoning on both textual and tabular data. Each question
in TabMWP is aligned with a tabular context, which is presented as an image,
semi-structured text, and a structured table. There are two types of questions:
free-text and multi-choice, and each problem is annotated with gold solutions
to reveal the multi-step reasoning process. We evaluate different pre-trained
models on TabMWP, including the GPT-3 model in a few-shot setting. As earlier
studies suggest, since few-shot GPT-3 relies on the selection of in-context
examples, its performance is unstable and can degrade to near chance. The
unstable issue is more severe when handling complex problems like TabMWP. To
mitigate this, we further propose a novel approach, PromptPG, which utilizes
policy gradient to learn to select in-context examples from a small amount of
training data and then constructs the corresponding prompt for the test
example. Experimental results show that our method outperforms the best
baseline by 5.31% on the accuracy metric and reduces the prediction variance
significantly compared to random selection, which verifies its effectiveness in
selecting in-context examples.
|
2209.15076 | Ho Hin Lee | Ho Hin Lee, Shunxing Bao, Yuankai Huo, Bennett A. Landman | 3D UX-Net: A Large Kernel Volumetric ConvNet Modernizing Hierarchical
Transformer for Medical Image Segmentation | Accepted to ICLR 2023 | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The recent 3D medical ViTs (e.g., SwinUNETR) achieve the state-of-the-art
performances on several 3D volumetric data benchmarks, including 3D medical
image segmentation. Hierarchical transformers (e.g., Swin Transformers)
reintroduced several ConvNet priors and further enhanced the practical
viability of adapting volumetric segmentation in 3D medical datasets. The
effectiveness of hybrid approaches is largely credited to the large receptive
field for non-local self-attention and the large number of model parameters. In
this work, we propose a lightweight volumetric ConvNet, termed 3D UX-Net, which
adapts the hierarchical transformer using ConvNet modules for robust volumetric
segmentation. Specifically, we revisit volumetric depth-wise convolutions with
large kernel size (e.g. starting from $7\times7\times7$) to enable the larger
global receptive fields, inspired by Swin Transformer. We further substitute
the multi-layer perceptron (MLP) in Swin Transformer blocks with pointwise
depth convolutions and enhance model performances with fewer normalization and
activation layers, thus reducing the number of model parameters. 3D UX-Net
competes favorably with current SOTA transformers (e.g. SwinUNETR) using three
challenging public datasets on volumetric brain and abdominal imaging: 1)
MICCAI Challenge 2021 FLARE, 2) MICCAI Challenge 2021 FeTA, and 3) MICCAI
Challenge 2022 AMOS. 3D UX-Net consistently outperforms SwinUNETR with
improvement from 0.929 to 0.938 Dice (FLARE2021) and 0.867 to 0.874 Dice
(Feta2021). We further evaluate the transfer learning capability of 3D UX-Net
with AMOS2022 and demonstrates another improvement of $2.27\%$ Dice (from 0.880
to 0.900). The source code with our proposed model are available at
https://github.com/MASILab/3DUX-Net.
| [
{
"version": "v1",
"created": "Thu, 29 Sep 2022 19:54:13 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Oct 2022 04:26:53 GMT"
},
{
"version": "v3",
"created": "Tue, 24 Jan 2023 04:55:36 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Mar 2023 03:58:57 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lee",
"Ho Hin",
""
],
[
"Bao",
"Shunxing",
""
],
[
"Huo",
"Yuankai",
""
],
[
"Landman",
"Bennett A.",
""
]
] | TITLE: 3D UX-Net: A Large Kernel Volumetric ConvNet Modernizing Hierarchical
Transformer for Medical Image Segmentation
ABSTRACT: The recent 3D medical ViTs (e.g., SwinUNETR) achieve the state-of-the-art
performances on several 3D volumetric data benchmarks, including 3D medical
image segmentation. Hierarchical transformers (e.g., Swin Transformers)
reintroduced several ConvNet priors and further enhanced the practical
viability of adapting volumetric segmentation in 3D medical datasets. The
effectiveness of hybrid approaches is largely credited to the large receptive
field for non-local self-attention and the large number of model parameters. In
this work, we propose a lightweight volumetric ConvNet, termed 3D UX-Net, which
adapts the hierarchical transformer using ConvNet modules for robust volumetric
segmentation. Specifically, we revisit volumetric depth-wise convolutions with
large kernel size (e.g. starting from $7\times7\times7$) to enable the larger
global receptive fields, inspired by Swin Transformer. We further substitute
the multi-layer perceptron (MLP) in Swin Transformer blocks with pointwise
depth convolutions and enhance model performances with fewer normalization and
activation layers, thus reducing the number of model parameters. 3D UX-Net
competes favorably with current SOTA transformers (e.g. SwinUNETR) using three
challenging public datasets on volumetric brain and abdominal imaging: 1)
MICCAI Challenge 2021 FLARE, 2) MICCAI Challenge 2021 FeTA, and 3) MICCAI
Challenge 2022 AMOS. 3D UX-Net consistently outperforms SwinUNETR with
improvement from 0.929 to 0.938 Dice (FLARE2021) and 0.867 to 0.874 Dice
(Feta2021). We further evaluate the transfer learning capability of 3D UX-Net
with AMOS2022 and demonstrates another improvement of $2.27\%$ Dice (from 0.880
to 0.900). The source code with our proposed model are available at
https://github.com/MASILab/3DUX-Net.
|
2210.01240 | Abulhair Saparov | Abulhair Saparov and He He | Language Models Are Greedy Reasoners: A Systematic Formal Analysis of
Chain-of-Thought | Published as a conference paper at ICLR 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have shown remarkable reasoning capabilities
given chain-of-thought prompts (examples with intermediate reasoning steps).
Existing benchmarks measure reasoning ability indirectly, by evaluating
accuracy on downstream tasks such as mathematical reasoning. However, it is
unclear how these models obtain the answers and whether they rely on simple
heuristics rather than the generated chain-of-thought. To enable systematic
exploration of the reasoning ability of LLMs, we present a new synthetic
question-answering dataset called PrOntoQA, where each example is generated
from a synthetic world model represented in first-order logic. This allows us
to parse the generated chain-of-thought into symbolic proofs for formal
analysis. Our analysis on InstructGPT and GPT-3 shows that LLMs are quite
capable of making correct individual deduction steps, and so are generally
capable of reasoning, even in fictional contexts. However, they have difficulty
with proof planning: When multiple valid deduction steps are available, they
are not able to systematically explore the different options.
| [
{
"version": "v1",
"created": "Mon, 3 Oct 2022 21:34:32 GMT"
},
{
"version": "v2",
"created": "Wed, 25 Jan 2023 05:33:23 GMT"
},
{
"version": "v3",
"created": "Thu, 26 Jan 2023 02:18:52 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Mar 2023 03:54:28 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Saparov",
"Abulhair",
""
],
[
"He",
"He",
""
]
] | TITLE: Language Models Are Greedy Reasoners: A Systematic Formal Analysis of
Chain-of-Thought
ABSTRACT: Large language models (LLMs) have shown remarkable reasoning capabilities
given chain-of-thought prompts (examples with intermediate reasoning steps).
Existing benchmarks measure reasoning ability indirectly, by evaluating
accuracy on downstream tasks such as mathematical reasoning. However, it is
unclear how these models obtain the answers and whether they rely on simple
heuristics rather than the generated chain-of-thought. To enable systematic
exploration of the reasoning ability of LLMs, we present a new synthetic
question-answering dataset called PrOntoQA, where each example is generated
from a synthetic world model represented in first-order logic. This allows us
to parse the generated chain-of-thought into symbolic proofs for formal
analysis. Our analysis on InstructGPT and GPT-3 shows that LLMs are quite
capable of making correct individual deduction steps, and so are generally
capable of reasoning, even in fictional contexts. However, they have difficulty
with proof planning: When multiple valid deduction steps are available, they
are not able to systematically explore the different options.
|
2210.03165 | Ali Shirali | Ali Shirali, Rediet Abebe, Moritz Hardt | A Theory of Dynamic Benchmarks | ICLR 2023 Version | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Dynamic benchmarks interweave model fitting and data collection in an attempt
to mitigate the limitations of static benchmarks. In contrast to an extensive
theoretical and empirical study of the static setting, the dynamic counterpart
lags behind due to limited empirical studies and no apparent theoretical
foundation to date. Responding to this deficit, we initiate a theoretical study
of dynamic benchmarking. We examine two realizations, one capturing current
practice and the other modeling more complex settings. In the first model,
where data collection and model fitting alternate sequentially, we prove that
model performance improves initially but can stall after only three rounds.
Label noise arising from, for instance, annotator disagreement leads to even
stronger negative results. Our second model generalizes the first to the case
where data collection and model fitting have a hierarchical dependency
structure. We show that this design guarantees strictly more progress than the
first, albeit at a significant increase in complexity. We support our
theoretical analysis by simulating dynamic benchmarks on two popular datasets.
These results illuminate the benefits and practical limitations of dynamic
benchmarking, providing both a theoretical foundation and a causal explanation
for observed bottlenecks in empirical work.
| [
{
"version": "v1",
"created": "Thu, 6 Oct 2022 18:56:46 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Oct 2022 19:19:18 GMT"
},
{
"version": "v3",
"created": "Wed, 1 Mar 2023 23:31:57 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shirali",
"Ali",
""
],
[
"Abebe",
"Rediet",
""
],
[
"Hardt",
"Moritz",
""
]
] | TITLE: A Theory of Dynamic Benchmarks
ABSTRACT: Dynamic benchmarks interweave model fitting and data collection in an attempt
to mitigate the limitations of static benchmarks. In contrast to an extensive
theoretical and empirical study of the static setting, the dynamic counterpart
lags behind due to limited empirical studies and no apparent theoretical
foundation to date. Responding to this deficit, we initiate a theoretical study
of dynamic benchmarking. We examine two realizations, one capturing current
practice and the other modeling more complex settings. In the first model,
where data collection and model fitting alternate sequentially, we prove that
model performance improves initially but can stall after only three rounds.
Label noise arising from, for instance, annotator disagreement leads to even
stronger negative results. Our second model generalizes the first to the case
where data collection and model fitting have a hierarchical dependency
structure. We show that this design guarantees strictly more progress than the
first, albeit at a significant increase in complexity. We support our
theoretical analysis by simulating dynamic benchmarks on two popular datasets.
These results illuminate the benefits and practical limitations of dynamic
benchmarking, providing both a theoretical foundation and a causal explanation
for observed bottlenecks in empirical work.
|
2210.03310 | Mengye Ren | Mengye Ren, Simon Kornblith, Renjie Liao, Geoffrey Hinton | Scaling Forward Gradient With Local Losses | 31 pages, ICLR 2023 | null | null | null | cs.LG cs.CV cs.NE | http://creativecommons.org/licenses/by/4.0/ | Forward gradient learning computes a noisy directional gradient and is a
biologically plausible alternative to backprop for learning deep neural
networks. However, the standard forward gradient algorithm, when applied
naively, suffers from high variance when the number of parameters to be learned
is large. In this paper, we propose a series of architectural and algorithmic
modifications that together make forward gradient learning practical for
standard deep learning benchmark tasks. We show that it is possible to
substantially reduce the variance of the forward gradient estimator by applying
perturbations to activations rather than weights. We further improve the
scalability of forward gradient by introducing a large number of local greedy
loss functions, each of which involves only a small number of learnable
parameters, and a new MLPMixer-inspired architecture, LocalMixer, that is more
suitable for local learning. Our approach matches backprop on MNIST and
CIFAR-10 and significantly outperforms previously proposed backprop-free
algorithms on ImageNet.
| [
{
"version": "v1",
"created": "Fri, 7 Oct 2022 03:52:27 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Feb 2023 02:15:53 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 03:08:10 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Ren",
"Mengye",
""
],
[
"Kornblith",
"Simon",
""
],
[
"Liao",
"Renjie",
""
],
[
"Hinton",
"Geoffrey",
""
]
] | TITLE: Scaling Forward Gradient With Local Losses
ABSTRACT: Forward gradient learning computes a noisy directional gradient and is a
biologically plausible alternative to backprop for learning deep neural
networks. However, the standard forward gradient algorithm, when applied
naively, suffers from high variance when the number of parameters to be learned
is large. In this paper, we propose a series of architectural and algorithmic
modifications that together make forward gradient learning practical for
standard deep learning benchmark tasks. We show that it is possible to
substantially reduce the variance of the forward gradient estimator by applying
perturbations to activations rather than weights. We further improve the
scalability of forward gradient by introducing a large number of local greedy
loss functions, each of which involves only a small number of learnable
parameters, and a new MLPMixer-inspired architecture, LocalMixer, that is more
suitable for local learning. Our approach matches backprop on MNIST and
CIFAR-10 and significantly outperforms previously proposed backprop-free
algorithms on ImageNet.
|
2210.06031 | Yuchong Sun | Yuchong Sun, Hongwei Xue, Ruihua Song, Bei Liu, Huan Yang, Jianlong Fu | Long-Form Video-Language Pre-Training with Multimodal Temporal
Contrastive Learning | Accepted by NeurIPS 2022 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large-scale video-language pre-training has shown significant improvement in
video-language understanding tasks. Previous studies of video-language
pretraining mainly focus on short-form videos (i.e., within 30 seconds) and
sentences, leaving long-form video-language pre-training rarely explored.
Directly learning representation from long-form videos and language may benefit
many long-form video-language understanding tasks. However, it is challenging
due to the difficulty of modeling long-range relationships and the heavy
computational burden caused by more frames. In this paper, we introduce a
Long-Form VIdeo-LAnguage pre-training model (LF-VILA) and train it on a
large-scale long-form video and paragraph dataset constructed from an existing
public dataset. To effectively capture the rich temporal dynamics and to better
align video and language in an efficient end-to-end manner, we introduce two
novel designs in our LF-VILA model. We first propose a Multimodal Temporal
Contrastive (MTC) loss to learn the temporal relation across different
modalities by encouraging fine-grained alignment between long-form videos and
paragraphs. Second, we propose a Hierarchical Temporal Window Attention (HTWA)
mechanism to effectively capture long-range dependency while reducing
computational cost in Transformer. We fine-tune the pre-trained LF-VILA model
on seven downstream long-form video-language understanding tasks of
paragraph-to-video retrieval and long-form video question-answering, and
achieve new state-of-the-art performances. Specifically, our model achieves
16.1% relative improvement on ActivityNet paragraph-to-video retrieval task and
2.4% on How2QA task, respectively. We release our code, dataset, and
pre-trained models at https://github.com/microsoft/XPretrain.
| [
{
"version": "v1",
"created": "Wed, 12 Oct 2022 09:08:27 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 09:05:43 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Sun",
"Yuchong",
""
],
[
"Xue",
"Hongwei",
""
],
[
"Song",
"Ruihua",
""
],
[
"Liu",
"Bei",
""
],
[
"Yang",
"Huan",
""
],
[
"Fu",
"Jianlong",
""
]
] | TITLE: Long-Form Video-Language Pre-Training with Multimodal Temporal
Contrastive Learning
ABSTRACT: Large-scale video-language pre-training has shown significant improvement in
video-language understanding tasks. Previous studies of video-language
pretraining mainly focus on short-form videos (i.e., within 30 seconds) and
sentences, leaving long-form video-language pre-training rarely explored.
Directly learning representation from long-form videos and language may benefit
many long-form video-language understanding tasks. However, it is challenging
due to the difficulty of modeling long-range relationships and the heavy
computational burden caused by more frames. In this paper, we introduce a
Long-Form VIdeo-LAnguage pre-training model (LF-VILA) and train it on a
large-scale long-form video and paragraph dataset constructed from an existing
public dataset. To effectively capture the rich temporal dynamics and to better
align video and language in an efficient end-to-end manner, we introduce two
novel designs in our LF-VILA model. We first propose a Multimodal Temporal
Contrastive (MTC) loss to learn the temporal relation across different
modalities by encouraging fine-grained alignment between long-form videos and
paragraphs. Second, we propose a Hierarchical Temporal Window Attention (HTWA)
mechanism to effectively capture long-range dependency while reducing
computational cost in Transformer. We fine-tune the pre-trained LF-VILA model
on seven downstream long-form video-language understanding tasks of
paragraph-to-video retrieval and long-form video question-answering, and
achieve new state-of-the-art performances. Specifically, our model achieves
16.1% relative improvement on ActivityNet paragraph-to-video retrieval task and
2.4% on How2QA task, respectively. We release our code, dataset, and
pre-trained models at https://github.com/microsoft/XPretrain.
|
2210.07998 | Sajad Movahedi | Sajad Movahedi, Melika Adabinejad, Ayyoob Imani, Arezou Keshavarz,
Mostafa Dehghani, Azadeh Shakery, Babak N. Araabi | $\Lambda$-DARTS: Mitigating Performance Collapse by Harmonizing
Operation Selection among Cells | Published as a conference paper at ICLR 2023 | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Differentiable neural architecture search (DARTS) is a popular method for
neural architecture search (NAS), which performs cell-search and utilizes
continuous relaxation to improve the search efficiency via gradient-based
optimization. The main shortcoming of DARTS is performance collapse, where the
discovered architecture suffers from a pattern of declining quality during
search. Performance collapse has become an important topic of research, with
many methods trying to solve the issue through either regularization or
fundamental changes to DARTS. However, the weight-sharing framework used for
cell-search in DARTS and the convergence of architecture parameters has not
been analyzed yet. In this paper, we provide a thorough and novel theoretical
and empirical analysis on DARTS and its point of convergence. We show that
DARTS suffers from a specific structural flaw due to its weight-sharing
framework that limits the convergence of DARTS to saturation points of the
softmax function. This point of convergence gives an unfair advantage to layers
closer to the output in choosing the optimal architecture, causing performance
collapse. We then propose two new regularization terms that aim to prevent
performance collapse by harmonizing operation selection via aligning gradients
of layers. Experimental results on six different search spaces and three
different datasets show that our method ($\Lambda$-DARTS) does indeed prevent
performance collapse, providing justification for our theoretical analysis and
the proposed remedy.
| [
{
"version": "v1",
"created": "Fri, 14 Oct 2022 17:54:01 GMT"
},
{
"version": "v2",
"created": "Wed, 1 Mar 2023 19:03:18 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Movahedi",
"Sajad",
""
],
[
"Adabinejad",
"Melika",
""
],
[
"Imani",
"Ayyoob",
""
],
[
"Keshavarz",
"Arezou",
""
],
[
"Dehghani",
"Mostafa",
""
],
[
"Shakery",
"Azadeh",
""
],
[
"Araabi",
"Babak N.",
""
]
] | TITLE: $\Lambda$-DARTS: Mitigating Performance Collapse by Harmonizing
Operation Selection among Cells
ABSTRACT: Differentiable neural architecture search (DARTS) is a popular method for
neural architecture search (NAS), which performs cell-search and utilizes
continuous relaxation to improve the search efficiency via gradient-based
optimization. The main shortcoming of DARTS is performance collapse, where the
discovered architecture suffers from a pattern of declining quality during
search. Performance collapse has become an important topic of research, with
many methods trying to solve the issue through either regularization or
fundamental changes to DARTS. However, the weight-sharing framework used for
cell-search in DARTS and the convergence of architecture parameters has not
been analyzed yet. In this paper, we provide a thorough and novel theoretical
and empirical analysis on DARTS and its point of convergence. We show that
DARTS suffers from a specific structural flaw due to its weight-sharing
framework that limits the convergence of DARTS to saturation points of the
softmax function. This point of convergence gives an unfair advantage to layers
closer to the output in choosing the optimal architecture, causing performance
collapse. We then propose two new regularization terms that aim to prevent
performance collapse by harmonizing operation selection via aligning gradients
of layers. Experimental results on six different search spaces and three
different datasets show that our method ($\Lambda$-DARTS) does indeed prevent
performance collapse, providing justification for our theoretical analysis and
the proposed remedy.
|
2210.12153 | Brandon Amos | Brandon Amos | On amortizing convex conjugates for optimal transport | ICLR 2023 | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | This paper focuses on computing the convex conjugate operation that arises
when solving Euclidean Wasserstein-2 optimal transport problems. This
conjugation, which is also referred to as the Legendre-Fenchel conjugate or
c-transform,is considered difficult to compute and in practice,Wasserstein-2
methods are limited by not being able to exactly conjugate the dual potentials
in continuous space. To overcome this, the computation of the conjugate can be
approximated with amortized optimization, which learns a model to predict the
conjugate. I show that combining amortized approximations to the conjugate with
a solver for fine-tuning significantly improves the quality of transport maps
learned for the Wasserstein-2 benchmark by Korotin et al. (2021a) and is able
to model many 2-dimensional couplings and flows considered in the literature.
All of the baselines, methods, and solvers in this paper are available at
http://github.com/facebookresearch/w2ot.
| [
{
"version": "v1",
"created": "Fri, 21 Oct 2022 17:59:24 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 01:21:27 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Amos",
"Brandon",
""
]
] | TITLE: On amortizing convex conjugates for optimal transport
ABSTRACT: This paper focuses on computing the convex conjugate operation that arises
when solving Euclidean Wasserstein-2 optimal transport problems. This
conjugation, which is also referred to as the Legendre-Fenchel conjugate or
c-transform,is considered difficult to compute and in practice,Wasserstein-2
methods are limited by not being able to exactly conjugate the dual potentials
in continuous space. To overcome this, the computation of the conjugate can be
approximated with amortized optimization, which learns a model to predict the
conjugate. I show that combining amortized approximations to the conjugate with
a solver for fine-tuning significantly improves the quality of transport maps
learned for the Wasserstein-2 benchmark by Korotin et al. (2021a) and is able
to model many 2-dimensional couplings and flows considered in the literature.
All of the baselines, methods, and solvers in this paper are available at
http://github.com/facebookresearch/w2ot.
|
2210.16142 | Yiqing Shen | Yiqing Shen, Baiyun Liu, Ruize Yu, Yudong Wang, Shaokang Wang,
Jiangfen Wu, Weidao Chen | Federated Learning for Chronic Obstructive Pulmonary Disease
Classification with Partial Personalized Attention Mechanism | 4 pages | null | 10.1109/BIBM55620.2022.9995355 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Chronic Obstructive Pulmonary Disease (COPD) is the fourth leading cause of
death worldwide. Yet, COPD diagnosis heavily relies on spirometric examination
as well as functional airway limitation, which may cause a considerable portion
of COPD patients underdiagnosed especially at the early stage. Recent advance
in deep learning (DL) has shown their promising potential in COPD
identification from CT images. However, with heterogeneous syndromes and
distinct phenotypes, DL models trained with CTs from one data center fail to
generalize on images from another center. Due to privacy regularizations, a
collaboration of distributed CT images into one centralized center is not
feasible. Federated learning (FL) approaches enable us to train with
distributed private data. Yet, routine FL solutions suffer from performance
degradation in the case where COPD CTs are not independent and identically
distributed (Non-IID). To address this issue, we propose a novel personalized
federated learning (PFL) method based on vision transformer (ViT) for
distributed and heterogeneous COPD CTs. To be more specific, we partially
personalize some heads in multiheaded self-attention layers to learn the
personalized attention for local data and retain the other heads shared to
extract the common attention. To the best of our knowledge, this is the first
proposal of a PFL framework specifically for ViT to identify COPD. Our
evaluation of a dataset set curated from six medical centers shows our method
outperforms the PFL approaches for convolutional neural networks.
| [
{
"version": "v1",
"created": "Fri, 28 Oct 2022 14:12:42 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shen",
"Yiqing",
""
],
[
"Liu",
"Baiyun",
""
],
[
"Yu",
"Ruize",
""
],
[
"Wang",
"Yudong",
""
],
[
"Wang",
"Shaokang",
""
],
[
"Wu",
"Jiangfen",
""
],
[
"Chen",
"Weidao",
""
]
] | TITLE: Federated Learning for Chronic Obstructive Pulmonary Disease
Classification with Partial Personalized Attention Mechanism
ABSTRACT: Chronic Obstructive Pulmonary Disease (COPD) is the fourth leading cause of
death worldwide. Yet, COPD diagnosis heavily relies on spirometric examination
as well as functional airway limitation, which may cause a considerable portion
of COPD patients underdiagnosed especially at the early stage. Recent advance
in deep learning (DL) has shown their promising potential in COPD
identification from CT images. However, with heterogeneous syndromes and
distinct phenotypes, DL models trained with CTs from one data center fail to
generalize on images from another center. Due to privacy regularizations, a
collaboration of distributed CT images into one centralized center is not
feasible. Federated learning (FL) approaches enable us to train with
distributed private data. Yet, routine FL solutions suffer from performance
degradation in the case where COPD CTs are not independent and identically
distributed (Non-IID). To address this issue, we propose a novel personalized
federated learning (PFL) method based on vision transformer (ViT) for
distributed and heterogeneous COPD CTs. To be more specific, we partially
personalize some heads in multiheaded self-attention layers to learn the
personalized attention for local data and retain the other heads shared to
extract the common attention. To the best of our knowledge, this is the first
proposal of a PFL framework specifically for ViT to identify COPD. Our
evaluation of a dataset set curated from six medical centers shows our method
outperforms the PFL approaches for convolutional neural networks.
|
2210.16966 | Pan Li | Siqi Miao, Yunan Luo, Mia Liu, Pan Li | Interpretable Geometric Deep Learning via Learnable Randomness Injection | ICLR 2023 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Point cloud data is ubiquitous in scientific fields. Recently, geometric deep
learning (GDL) has been widely applied to solve prediction tasks with such
data. However, GDL models are often complicated and hardly interpretable, which
poses concerns to scientists who are to deploy these models in scientific
analysis and experiments. This work proposes a general mechanism, learnable
randomness injection (LRI), which allows building inherently interpretable
models based on general GDL backbones. LRI-induced models, once trained, can
detect the points in the point cloud data that carry information indicative of
the prediction label. We also propose four datasets from real scientific
applications that cover the domains of high-energy physics and biochemistry to
evaluate the LRI mechanism. Compared with previous post-hoc interpretation
methods, the points detected by LRI align much better and stabler with the
ground-truth patterns that have actual scientific meanings. LRI is grounded by
the information bottleneck principle, and thus LRI-induced models are also more
robust to distribution shifts between training and test scenarios. Our code and
datasets are available at \url{https://github.com/Graph-COM/LRI}.
| [
{
"version": "v1",
"created": "Sun, 30 Oct 2022 22:24:43 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 17:21:16 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Miao",
"Siqi",
""
],
[
"Luo",
"Yunan",
""
],
[
"Liu",
"Mia",
""
],
[
"Li",
"Pan",
""
]
] | TITLE: Interpretable Geometric Deep Learning via Learnable Randomness Injection
ABSTRACT: Point cloud data is ubiquitous in scientific fields. Recently, geometric deep
learning (GDL) has been widely applied to solve prediction tasks with such
data. However, GDL models are often complicated and hardly interpretable, which
poses concerns to scientists who are to deploy these models in scientific
analysis and experiments. This work proposes a general mechanism, learnable
randomness injection (LRI), which allows building inherently interpretable
models based on general GDL backbones. LRI-induced models, once trained, can
detect the points in the point cloud data that carry information indicative of
the prediction label. We also propose four datasets from real scientific
applications that cover the domains of high-energy physics and biochemistry to
evaluate the LRI mechanism. Compared with previous post-hoc interpretation
methods, the points detected by LRI align much better and stabler with the
ground-truth patterns that have actual scientific meanings. LRI is grounded by
the information bottleneck principle, and thus LRI-induced models are also more
robust to distribution shifts between training and test scenarios. Our code and
datasets are available at \url{https://github.com/Graph-COM/LRI}.
|
2211.00854 | Tatiana Likhomanenko | Dan Berrebbi, Ronan Collobert, Navdeep Jaitly, Tatiana Likhomanenko | More Speaking or More Speakers? | ICASSP 2023 | null | null | null | cs.LG cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-training (ST) and self-supervised learning (SSL) methods have
demonstrated strong improvements in automatic speech recognition (ASR). In
spite of these advances, to the best of our knowledge, there is no analysis of
how the composition of the labelled and unlabelled datasets used in these
methods affects the results. In this work we aim to analyse the effect of
number of speakers in the training data on a recent SSL algorithm (wav2vec
2.0), and a recent ST algorithm (slimIPL). We perform a systematic analysis on
both labeled and unlabeled data by varying the number of speakers while keeping
the number of hours fixed and vice versa. Our findings suggest that SSL
requires a large amount of unlabeled data to produce high accuracy results,
while ST requires a sufficient number of speakers in the labelled data,
especially in the low-regime setting. In this manner these two approaches
improve supervised learning in different regimes of data composition.
| [
{
"version": "v1",
"created": "Wed, 2 Nov 2022 03:50:40 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 06:43:56 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Berrebbi",
"Dan",
""
],
[
"Collobert",
"Ronan",
""
],
[
"Jaitly",
"Navdeep",
""
],
[
"Likhomanenko",
"Tatiana",
""
]
] | TITLE: More Speaking or More Speakers?
ABSTRACT: Self-training (ST) and self-supervised learning (SSL) methods have
demonstrated strong improvements in automatic speech recognition (ASR). In
spite of these advances, to the best of our knowledge, there is no analysis of
how the composition of the labelled and unlabelled datasets used in these
methods affects the results. In this work we aim to analyse the effect of
number of speakers in the training data on a recent SSL algorithm (wav2vec
2.0), and a recent ST algorithm (slimIPL). We perform a systematic analysis on
both labeled and unlabeled data by varying the number of speakers while keeping
the number of hours fixed and vice versa. Our findings suggest that SSL
requires a large amount of unlabeled data to produce high accuracy results,
while ST requires a sufficient number of speakers in the labelled data,
especially in the low-regime setting. In this manner these two approaches
improve supervised learning in different regimes of data composition.
|
2211.02432 | Chen-Chou Lo | Chen-Chou Lo and Patrick Vandewalle | RCDPT: Radar-Camera fusion Dense Prediction Transformer | 5 pages, 2 figures and 1 table, accepted to ICASSP2023 | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, transformer networks have outperformed traditional deep neural
networks in natural language processing and show a large potential in many
computer vision tasks compared to convolutional backbones. In the original
transformer, readout tokens are used as designated vectors for aggregating
information from other tokens. However, the performance of using readout tokens
in a vision transformer is limited. Therefore, we propose a novel fusion
strategy to integrate radar data into a dense prediction transformer network by
reassembling camera representations with radar representations. Instead of
using readout tokens, radar representations contribute additional depth
information to a monocular depth estimation model and improve performance. We
further investigate different fusion approaches that are commonly used for
integrating additional modality in a dense prediction transformer network. The
experiments are conducted on the nuScenes dataset, which includes camera
images, lidar, and radar data. The results show that our proposed method yields
better performance than the commonly used fusion strategies and outperforms
existing convolutional depth estimation models that fuse camera images and
radar.
| [
{
"version": "v1",
"created": "Fri, 4 Nov 2022 13:16:20 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 15:00:46 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lo",
"Chen-Chou",
""
],
[
"Vandewalle",
"Patrick",
""
]
] | TITLE: RCDPT: Radar-Camera fusion Dense Prediction Transformer
ABSTRACT: Recently, transformer networks have outperformed traditional deep neural
networks in natural language processing and show a large potential in many
computer vision tasks compared to convolutional backbones. In the original
transformer, readout tokens are used as designated vectors for aggregating
information from other tokens. However, the performance of using readout tokens
in a vision transformer is limited. Therefore, we propose a novel fusion
strategy to integrate radar data into a dense prediction transformer network by
reassembling camera representations with radar representations. Instead of
using readout tokens, radar representations contribute additional depth
information to a monocular depth estimation model and improve performance. We
further investigate different fusion approaches that are commonly used for
integrating additional modality in a dense prediction transformer network. The
experiments are conducted on the nuScenes dataset, which includes camera
images, lidar, and radar data. The results show that our proposed method yields
better performance than the commonly used fusion strategies and outperforms
existing convolutional depth estimation models that fuse camera images and
radar.
|
2211.08458 | Leo Feng | Leo Feng, Hossein Hajimirsadeghi, Yoshua Bengio, Mohamed Osama Ahmed | Latent Bottlenecked Attentive Neural Processes | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Neural Processes (NPs) are popular methods in meta-learning that can estimate
predictive uncertainty on target datapoints by conditioning on a context
dataset. Previous state-of-the-art method Transformer Neural Processes (TNPs)
achieve strong performance but require quadratic computation with respect to
the number of context datapoints, significantly limiting its scalability.
Conversely, existing sub-quadratic NP variants perform significantly worse than
that of TNPs. Tackling this issue, we propose Latent Bottlenecked Attentive
Neural Processes (LBANPs), a new computationally efficient sub-quadratic NP
variant, that has a querying computational complexity independent of the number
of context datapoints. The model encodes the context dataset into a constant
number of latent vectors on which self-attention is performed. When making
predictions, the model retrieves higher-order information from the context
dataset via multiple cross-attention mechanisms on the latent vectors. We
empirically show that LBANPs achieve results competitive with the
state-of-the-art on meta-regression, image completion, and contextual
multi-armed bandits. We demonstrate that LBANPs can trade-off the computational
cost and performance according to the number of latent vectors. Finally, we
show LBANPs can scale beyond existing attention-based NP variants to larger
dataset settings.
| [
{
"version": "v1",
"created": "Tue, 15 Nov 2022 19:21:41 GMT"
},
{
"version": "v2",
"created": "Fri, 24 Feb 2023 23:24:29 GMT"
},
{
"version": "v3",
"created": "Wed, 1 Mar 2023 20:34:17 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Feng",
"Leo",
""
],
[
"Hajimirsadeghi",
"Hossein",
""
],
[
"Bengio",
"Yoshua",
""
],
[
"Ahmed",
"Mohamed Osama",
""
]
] | TITLE: Latent Bottlenecked Attentive Neural Processes
ABSTRACT: Neural Processes (NPs) are popular methods in meta-learning that can estimate
predictive uncertainty on target datapoints by conditioning on a context
dataset. Previous state-of-the-art method Transformer Neural Processes (TNPs)
achieve strong performance but require quadratic computation with respect to
the number of context datapoints, significantly limiting its scalability.
Conversely, existing sub-quadratic NP variants perform significantly worse than
that of TNPs. Tackling this issue, we propose Latent Bottlenecked Attentive
Neural Processes (LBANPs), a new computationally efficient sub-quadratic NP
variant, that has a querying computational complexity independent of the number
of context datapoints. The model encodes the context dataset into a constant
number of latent vectors on which self-attention is performed. When making
predictions, the model retrieves higher-order information from the context
dataset via multiple cross-attention mechanisms on the latent vectors. We
empirically show that LBANPs achieve results competitive with the
state-of-the-art on meta-regression, image completion, and contextual
multi-armed bandits. We demonstrate that LBANPs can trade-off the computational
cost and performance according to the number of latent vectors. Finally, we
show LBANPs can scale beyond existing attention-based NP variants to larger
dataset settings.
|
2211.12732 | Joshua Knights Mr | Joshua Knights, Kavisha Vidanapathirana, Milad Ramezani, Sridha
Sridharan, Clinton Fookes, Peyman Moghadam | Wild-Places: A Large-Scale Dataset for Lidar Place Recognition in
Unstructured Natural Environments | Equal Contribution from first two authors Accepted to ICRA2023
Website link: https://csiro-robotics.github.io/Wild-Places/ | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many existing datasets for lidar place recognition are solely representative
of structured urban environments, and have recently been saturated in
performance by deep learning based approaches. Natural and unstructured
environments present many additional challenges for the tasks of long-term
localisation but these environments are not represented in currently available
datasets. To address this we introduce Wild-Places, a challenging large-scale
dataset for lidar place recognition in unstructured, natural environments.
Wild-Places contains eight lidar sequences collected with a handheld sensor
payload over the course of fourteen months, containing a total of 63K
undistorted lidar submaps along with accurate 6DoF ground truth. Our dataset
contains multiple revisits both within and between sequences, allowing for both
intra-sequence (i.e. loop closure detection) and inter-sequence (i.e.
re-localisation) place recognition. We also benchmark several state-of-the-art
approaches to demonstrate the challenges that this dataset introduces,
particularly the case of long-term place recognition due to natural
environments changing over time. Our dataset and code will be available at
https://csiro-robotics.github.io/Wild-Places.
| [
{
"version": "v1",
"created": "Wed, 23 Nov 2022 06:50:31 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Nov 2022 07:17:10 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 06:45:21 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Knights",
"Joshua",
""
],
[
"Vidanapathirana",
"Kavisha",
""
],
[
"Ramezani",
"Milad",
""
],
[
"Sridharan",
"Sridha",
""
],
[
"Fookes",
"Clinton",
""
],
[
"Moghadam",
"Peyman",
""
]
] | TITLE: Wild-Places: A Large-Scale Dataset for Lidar Place Recognition in
Unstructured Natural Environments
ABSTRACT: Many existing datasets for lidar place recognition are solely representative
of structured urban environments, and have recently been saturated in
performance by deep learning based approaches. Natural and unstructured
environments present many additional challenges for the tasks of long-term
localisation but these environments are not represented in currently available
datasets. To address this we introduce Wild-Places, a challenging large-scale
dataset for lidar place recognition in unstructured, natural environments.
Wild-Places contains eight lidar sequences collected with a handheld sensor
payload over the course of fourteen months, containing a total of 63K
undistorted lidar submaps along with accurate 6DoF ground truth. Our dataset
contains multiple revisits both within and between sequences, allowing for both
intra-sequence (i.e. loop closure detection) and inter-sequence (i.e.
re-localisation) place recognition. We also benchmark several state-of-the-art
approaches to demonstrate the challenges that this dataset introduces,
particularly the case of long-term place recognition due to natural
environments changing over time. Our dataset and code will be available at
https://csiro-robotics.github.io/Wild-Places.
|
2211.13956 | Khaled Koutini | Khaled Koutini, Shahed Masoudian, Florian Schmid, Hamid Eghbal-zadeh,
Jan Schl\"uter, Gerhard Widmer | Learning General Audio Representations with Large-Scale Training of
Patchout Audio Transformers | will apear in HEAR: Holistic Evaluation of Audio Representations
Proceedings of Machine Learning Research PMLR 166. Source code:
https://github.com/kkoutini/passt_hear21 | Proceedings of Machine Learning Research v166 (2022) 65-89 | null | null | cs.SD cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The success of supervised deep learning methods is largely due to their
ability to learn relevant features from raw data. Deep Neural Networks (DNNs)
trained on large-scale datasets are capable of capturing a diverse set of
features, and learning a representation that can generalize onto unseen tasks
and datasets that are from the same domain. Hence, these models can be used as
powerful feature extractors, in combination with shallower models as
classifiers, for smaller tasks and datasets where the amount of training data
is insufficient for learning an end-to-end model from scratch. During the past
years, Convolutional Neural Networks (CNNs) have largely been the method of
choice for audio processing. However, recently attention-based transformer
models have demonstrated great potential in supervised settings, outperforming
CNNs. In this work, we investigate the use of audio transformers trained on
large-scale datasets to learn general-purpose representations. We study how the
different setups in these audio transformers affect the quality of their
embeddings. We experiment with the models' time resolution, extracted embedding
level, and receptive fields in order to see how they affect performance on a
variety of tasks and datasets, following the HEAR 2021 NeurIPS challenge
evaluation setup. Our results show that representations extracted by audio
transformers outperform CNN representations. Furthermore, we will show that
transformers trained on Audioset can be extremely effective representation
extractors for a wide range of downstream tasks.
| [
{
"version": "v1",
"created": "Fri, 25 Nov 2022 08:39:12 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 10:53:40 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Koutini",
"Khaled",
""
],
[
"Masoudian",
"Shahed",
""
],
[
"Schmid",
"Florian",
""
],
[
"Eghbal-zadeh",
"Hamid",
""
],
[
"Schlüter",
"Jan",
""
],
[
"Widmer",
"Gerhard",
""
]
] | TITLE: Learning General Audio Representations with Large-Scale Training of
Patchout Audio Transformers
ABSTRACT: The success of supervised deep learning methods is largely due to their
ability to learn relevant features from raw data. Deep Neural Networks (DNNs)
trained on large-scale datasets are capable of capturing a diverse set of
features, and learning a representation that can generalize onto unseen tasks
and datasets that are from the same domain. Hence, these models can be used as
powerful feature extractors, in combination with shallower models as
classifiers, for smaller tasks and datasets where the amount of training data
is insufficient for learning an end-to-end model from scratch. During the past
years, Convolutional Neural Networks (CNNs) have largely been the method of
choice for audio processing. However, recently attention-based transformer
models have demonstrated great potential in supervised settings, outperforming
CNNs. In this work, we investigate the use of audio transformers trained on
large-scale datasets to learn general-purpose representations. We study how the
different setups in these audio transformers affect the quality of their
embeddings. We experiment with the models' time resolution, extracted embedding
level, and receptive fields in order to see how they affect performance on a
variety of tasks and datasets, following the HEAR 2021 NeurIPS challenge
evaluation setup. Our results show that representations extracted by audio
transformers outperform CNN representations. Furthermore, we will show that
transformers trained on Audioset can be extremely effective representation
extractors for a wide range of downstream tasks.
|
2212.02295 | Yeonguk Yu | Yeonguk Yu, Sungho Shin, Seongju Lee, Changhyun Jun, Kyoobin Lee | Block Selection Method for Using Feature Norm in Out-of-distribution
Detection | CVPR2023 accepted; Code is available in
https://github.com/gist-ailab/block-selection-for-OOD-detection | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Detecting out-of-distribution (OOD) inputs during the inference stage is
crucial for deploying neural networks in the real world. Previous methods
commonly relied on the output of a network derived from the highly activated
feature map. In this study, we first revealed that a norm of the feature map
obtained from the other block than the last block can be a better indicator of
OOD detection. Motivated by this, we propose a simple framework consisting of
FeatureNorm: a norm of the feature map and NormRatio: a ratio of FeatureNorm
for ID and OOD to measure the OOD detection performance of each block. In
particular, to select the block that provides the largest difference between
FeatureNorm of ID and FeatureNorm of OOD, we create Jigsaw puzzle images as
pseudo OOD from ID training samples and calculate NormRatio, and the block with
the largest value is selected. After the suitable block is selected, OOD
detection with the FeatureNorm outperforms other OOD detection methods by
reducing FPR95 by up to 52.77% on CIFAR10 benchmark and by up to 48.53% on
ImageNet benchmark. We demonstrate that our framework can generalize to various
architectures and the importance of block selection, which can improve previous
OOD detection methods as well.
| [
{
"version": "v1",
"created": "Mon, 5 Dec 2022 14:19:21 GMT"
},
{
"version": "v2",
"created": "Sat, 10 Dec 2022 07:42:59 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 15:23:11 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yu",
"Yeonguk",
""
],
[
"Shin",
"Sungho",
""
],
[
"Lee",
"Seongju",
""
],
[
"Jun",
"Changhyun",
""
],
[
"Lee",
"Kyoobin",
""
]
] | TITLE: Block Selection Method for Using Feature Norm in Out-of-distribution
Detection
ABSTRACT: Detecting out-of-distribution (OOD) inputs during the inference stage is
crucial for deploying neural networks in the real world. Previous methods
commonly relied on the output of a network derived from the highly activated
feature map. In this study, we first revealed that a norm of the feature map
obtained from the other block than the last block can be a better indicator of
OOD detection. Motivated by this, we propose a simple framework consisting of
FeatureNorm: a norm of the feature map and NormRatio: a ratio of FeatureNorm
for ID and OOD to measure the OOD detection performance of each block. In
particular, to select the block that provides the largest difference between
FeatureNorm of ID and FeatureNorm of OOD, we create Jigsaw puzzle images as
pseudo OOD from ID training samples and calculate NormRatio, and the block with
the largest value is selected. After the suitable block is selected, OOD
detection with the FeatureNorm outperforms other OOD detection methods by
reducing FPR95 by up to 52.77% on CIFAR10 benchmark and by up to 48.53% on
ImageNet benchmark. We demonstrate that our framework can generalize to various
architectures and the importance of block selection, which can improve previous
OOD detection methods as well.
|
2212.03125 | Boxi Wu | Minghao Chen, Renbo Tu, Chenxi Huang, Yuqi Lin, Boxi Wu, Deng Cai | Self-supervised and Weakly Supervised Contrastive Learning for
Frame-wise Action Representations | author conflicts | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Previous work on action representation learning focused on global
representations for short video clips. In contrast, many practical
applications, such as video alignment, strongly demand learning the intensive
representation of long videos. In this paper, we introduce a new framework of
contrastive action representation learning (CARL) to learn frame-wise action
representation in a self-supervised or weakly-supervised manner, especially for
long videos. Specifically, we introduce a simple but effective video encoder
that considers both spatial and temporal context by combining convolution and
transformer. Inspired by the recent massive progress in self-supervised
learning, we propose a new sequence contrast loss (SCL) applied to two related
views obtained by expanding a series of spatio-temporal data in two versions.
One is the self-supervised version that optimizes embedding space by minimizing
KL-divergence between sequence similarity of two augmented views and prior
Gaussian distribution of timestamp distance. The other is the weakly-supervised
version that builds more sample pairs among videos using video-level labels by
dynamic time wrapping (DTW). Experiments on FineGym, PennAction, and Pouring
datasets show that our method outperforms previous state-of-the-art by a large
margin for downstream fine-grained action classification and even faster
inference. Surprisingly, although without training on paired videos like in
previous works, our self-supervised version also shows outstanding performance
in video alignment and fine-grained frame retrieval tasks.
| [
{
"version": "v1",
"created": "Tue, 6 Dec 2022 16:42:22 GMT"
},
{
"version": "v2",
"created": "Sun, 11 Dec 2022 04:58:51 GMT"
},
{
"version": "v3",
"created": "Fri, 23 Dec 2022 13:10:38 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Mar 2023 04:44:43 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chen",
"Minghao",
""
],
[
"Tu",
"Renbo",
""
],
[
"Huang",
"Chenxi",
""
],
[
"Lin",
"Yuqi",
""
],
[
"Wu",
"Boxi",
""
],
[
"Cai",
"Deng",
""
]
] | TITLE: Self-supervised and Weakly Supervised Contrastive Learning for
Frame-wise Action Representations
ABSTRACT: Previous work on action representation learning focused on global
representations for short video clips. In contrast, many practical
applications, such as video alignment, strongly demand learning the intensive
representation of long videos. In this paper, we introduce a new framework of
contrastive action representation learning (CARL) to learn frame-wise action
representation in a self-supervised or weakly-supervised manner, especially for
long videos. Specifically, we introduce a simple but effective video encoder
that considers both spatial and temporal context by combining convolution and
transformer. Inspired by the recent massive progress in self-supervised
learning, we propose a new sequence contrast loss (SCL) applied to two related
views obtained by expanding a series of spatio-temporal data in two versions.
One is the self-supervised version that optimizes embedding space by minimizing
KL-divergence between sequence similarity of two augmented views and prior
Gaussian distribution of timestamp distance. The other is the weakly-supervised
version that builds more sample pairs among videos using video-level labels by
dynamic time wrapping (DTW). Experiments on FineGym, PennAction, and Pouring
datasets show that our method outperforms previous state-of-the-art by a large
margin for downstream fine-grained action classification and even faster
inference. Surprisingly, although without training on paired videos like in
previous works, our self-supervised version also shows outstanding performance
in video alignment and fine-grained frame retrieval tasks.
|
2212.08570 | Harry Coppock Mr | Harry Coppock, George Nicholson, Ivan Kiskin, Vasiliki Koutra, Kieran
Baker, Jobie Budd, Richard Payne, Emma Karoune, David Hurley, Alexander
Titcomb, Sabrina Egglestone, Ana Tendero Ca\~nadas, Lorraine Butler, Radka
Jersakova, Jonathon Mellor, Selina Patel, Tracey Thornley, Peter Diggle,
Sylvia Richardson, Josef Packham, Bj\"orn W. Schuller, Davide Pigoli, Steven
Gilmour, Stephen Roberts, Chris Holmes | Audio-based AI classifiers show no evidence of improved COVID-19
screening over simple symptoms checkers | null | null | null | null | cs.SD cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | Recent work has reported that AI classifiers trained on audio recordings can
accurately predict severe acute respiratory syndrome coronavirus 2 (SARSCoV2)
infection status. Here, we undertake a large scale study of audio-based deep
learning classifiers, as part of the UK governments pandemic response. We
collect and analyse a dataset of audio recordings from 67,842 individuals with
linked metadata, including reverse transcription polymerase chain reaction
(PCR) test outcomes, of whom 23,514 tested positive for SARS CoV 2. Subjects
were recruited via the UK governments National Health Service Test-and-Trace
programme and the REal-time Assessment of Community Transmission (REACT)
randomised surveillance survey. In an unadjusted analysis of our dataset AI
classifiers predict SARS-CoV-2 infection status with high accuracy (Receiver
Operating Characteristic Area Under the Curve (ROCAUC) 0.846 [0.838, 0.854])
consistent with the findings of previous studies. However, after matching on
measured confounders, such as age, gender, and self reported symptoms, our
classifiers performance is much weaker (ROC-AUC 0.619 [0.594, 0.644]). Upon
quantifying the utility of audio based classifiers in practical settings, we
find them to be outperformed by simple predictive scores based on user reported
symptoms.
| [
{
"version": "v1",
"created": "Thu, 15 Dec 2022 15:44:02 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 18:12:11 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Coppock",
"Harry",
""
],
[
"Nicholson",
"George",
""
],
[
"Kiskin",
"Ivan",
""
],
[
"Koutra",
"Vasiliki",
""
],
[
"Baker",
"Kieran",
""
],
[
"Budd",
"Jobie",
""
],
[
"Payne",
"Richard",
""
],
[
"Karoune",
"Emma",
""
],
[
"Hurley",
"David",
""
],
[
"Titcomb",
"Alexander",
""
],
[
"Egglestone",
"Sabrina",
""
],
[
"Cañadas",
"Ana Tendero",
""
],
[
"Butler",
"Lorraine",
""
],
[
"Jersakova",
"Radka",
""
],
[
"Mellor",
"Jonathon",
""
],
[
"Patel",
"Selina",
""
],
[
"Thornley",
"Tracey",
""
],
[
"Diggle",
"Peter",
""
],
[
"Richardson",
"Sylvia",
""
],
[
"Packham",
"Josef",
""
],
[
"Schuller",
"Björn W.",
""
],
[
"Pigoli",
"Davide",
""
],
[
"Gilmour",
"Steven",
""
],
[
"Roberts",
"Stephen",
""
],
[
"Holmes",
"Chris",
""
]
] | TITLE: Audio-based AI classifiers show no evidence of improved COVID-19
screening over simple symptoms checkers
ABSTRACT: Recent work has reported that AI classifiers trained on audio recordings can
accurately predict severe acute respiratory syndrome coronavirus 2 (SARSCoV2)
infection status. Here, we undertake a large scale study of audio-based deep
learning classifiers, as part of the UK governments pandemic response. We
collect and analyse a dataset of audio recordings from 67,842 individuals with
linked metadata, including reverse transcription polymerase chain reaction
(PCR) test outcomes, of whom 23,514 tested positive for SARS CoV 2. Subjects
were recruited via the UK governments National Health Service Test-and-Trace
programme and the REal-time Assessment of Community Transmission (REACT)
randomised surveillance survey. In an unadjusted analysis of our dataset AI
classifiers predict SARS-CoV-2 infection status with high accuracy (Receiver
Operating Characteristic Area Under the Curve (ROCAUC) 0.846 [0.838, 0.854])
consistent with the findings of previous studies. However, after matching on
measured confounders, such as age, gender, and self reported symptoms, our
classifiers performance is much weaker (ROC-AUC 0.619 [0.594, 0.644]). Upon
quantifying the utility of audio based classifiers in practical settings, we
find them to be outperformed by simple predictive scores based on user reported
symptoms.
|
2212.09748 | William Peebles | William Peebles, Saining Xie | Scalable Diffusion Models with Transformers | Code, project page and videos available at
https://www.wpeebles.com/DiT | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | We explore a new class of diffusion models based on the transformer
architecture. We train latent diffusion models of images, replacing the
commonly-used U-Net backbone with a transformer that operates on latent
patches. We analyze the scalability of our Diffusion Transformers (DiTs)
through the lens of forward pass complexity as measured by Gflops. We find that
DiTs with higher Gflops -- through increased transformer depth/width or
increased number of input tokens -- consistently have lower FID. In addition to
possessing good scalability properties, our largest DiT-XL/2 models outperform
all prior diffusion models on the class-conditional ImageNet 512x512 and
256x256 benchmarks, achieving a state-of-the-art FID of 2.27 on the latter.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 18:59:58 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 09:06:55 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Peebles",
"William",
""
],
[
"Xie",
"Saining",
""
]
] | TITLE: Scalable Diffusion Models with Transformers
ABSTRACT: We explore a new class of diffusion models based on the transformer
architecture. We train latent diffusion models of images, replacing the
commonly-used U-Net backbone with a transformer that operates on latent
patches. We analyze the scalability of our Diffusion Transformers (DiTs)
through the lens of forward pass complexity as measured by Gflops. We find that
DiTs with higher Gflops -- through increased transformer depth/width or
increased number of input tokens -- consistently have lower FID. In addition to
possessing good scalability properties, our largest DiT-XL/2 models outperform
all prior diffusion models on the class-conditional ImageNet 512x512 and
256x256 benchmarks, achieving a state-of-the-art FID of 2.27 on the latter.
|
2301.03982 | Mohak Chadha | Mohak Chadha, Nils Krueger, Jophin John, Anshul Jindal, Michael
Gerndt, Shajulin Benedict | Exploring the Use of WebAssembly in HPC | ACM SIGPLAN PPoPP 2023 | null | 10.1145/3572848.3577436 | null | cs.DC cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Containerization approaches based on namespaces offered by the Linux kernel
have seen an increasing popularity in the HPC community both as a means to
isolate applications and as a format to package and distribute them. However,
their adoption and usage in HPC systems faces several challenges. These include
difficulties in unprivileged running and building of scientific application
container images directly on HPC resources, increasing heterogeneity of HPC
architectures, and access to specialized networking libraries available only on
HPC systems. These challenges of container-based HPC application development
closely align with the several advantages that a new universal intermediate
binary format called WebAssembly (Wasm) has to offer. These include a
lightweight userspace isolation mechanism and portability across operating
systems and processor architectures. In this paper, we explore the usage of
Wasm as a distribution format for MPI-based HPC applications. To this end, we
present MPIWasm, a novel Wasm embedder for MPI-based HPC applications that
enables high-performance execution of Wasm code, has low-overhead for MPI
calls, and supports high-performance networking interconnects present on HPC
systems. We evaluate the performance and overhead of MPIWasm on a production
HPC system and AWS Graviton2 nodes using standardized HPC benchmarks. Results
from our experiments demonstrate that MPIWasm delivers competitive native
application performance across all scenarios. Moreover, we observe that Wasm
binaries are 139.5x smaller on average as compared to the statically-linked
binaries for the different standardized benchmarks.
| [
{
"version": "v1",
"created": "Tue, 10 Jan 2023 14:25:00 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chadha",
"Mohak",
""
],
[
"Krueger",
"Nils",
""
],
[
"John",
"Jophin",
""
],
[
"Jindal",
"Anshul",
""
],
[
"Gerndt",
"Michael",
""
],
[
"Benedict",
"Shajulin",
""
]
] | TITLE: Exploring the Use of WebAssembly in HPC
ABSTRACT: Containerization approaches based on namespaces offered by the Linux kernel
have seen an increasing popularity in the HPC community both as a means to
isolate applications and as a format to package and distribute them. However,
their adoption and usage in HPC systems faces several challenges. These include
difficulties in unprivileged running and building of scientific application
container images directly on HPC resources, increasing heterogeneity of HPC
architectures, and access to specialized networking libraries available only on
HPC systems. These challenges of container-based HPC application development
closely align with the several advantages that a new universal intermediate
binary format called WebAssembly (Wasm) has to offer. These include a
lightweight userspace isolation mechanism and portability across operating
systems and processor architectures. In this paper, we explore the usage of
Wasm as a distribution format for MPI-based HPC applications. To this end, we
present MPIWasm, a novel Wasm embedder for MPI-based HPC applications that
enables high-performance execution of Wasm code, has low-overhead for MPI
calls, and supports high-performance networking interconnects present on HPC
systems. We evaluate the performance and overhead of MPIWasm on a production
HPC system and AWS Graviton2 nodes using standardized HPC benchmarks. Results
from our experiments demonstrate that MPIWasm delivers competitive native
application performance across all scenarios. Moreover, we observe that Wasm
binaries are 139.5x smaller on average as compared to the statically-linked
binaries for the different standardized benchmarks.
|
2301.12112 | Danqing Wang | Danqing Wang, Fei Ye, Hao Zhou | On Pre-trained Language Models for Antibody | Accepted in ICLR 2023 | null | null | null | cs.CL q-bio.BM | http://creativecommons.org/licenses/by/4.0/ | Antibodies are vital proteins offering robust protection for the human body
from pathogens. The development of general protein and antibody-specific
pre-trained language models both facilitate antibody prediction tasks. However,
there have been limited studies that comprehensively explore the representation
capability of distinct pre-trained language models on different antibody tasks.
To investigate the problem, we aim to answer several key questions in this
paper, such as how pre-trained language models perform in antibody tasks with
different specificity and how introducing specific biological mechanisms to the
pre-training process can benefit the model. Additionally, we evaluate if the
learned antibody pre-trained representations can be applied to real-world
antibody problems, like drug discovery and immune process understanding.
Previously, no benchmark available largely hindered the study to answer these
questions. To aid in our investigation, we provide an AnTibody Understanding
Evaluation (ATUE) benchmark. We comprehensively evaluate the performance of
protein pre-trained language models by empirical study along with conclusions
and new insights. Our ATUE and code are released at
https://github.com/dqwang122/EATLM.
| [
{
"version": "v1",
"created": "Sat, 28 Jan 2023 07:05:15 GMT"
},
{
"version": "v2",
"created": "Wed, 1 Mar 2023 20:18:25 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Wang",
"Danqing",
""
],
[
"Ye",
"Fei",
""
],
[
"Zhou",
"Hao",
""
]
] | TITLE: On Pre-trained Language Models for Antibody
ABSTRACT: Antibodies are vital proteins offering robust protection for the human body
from pathogens. The development of general protein and antibody-specific
pre-trained language models both facilitate antibody prediction tasks. However,
there have been limited studies that comprehensively explore the representation
capability of distinct pre-trained language models on different antibody tasks.
To investigate the problem, we aim to answer several key questions in this
paper, such as how pre-trained language models perform in antibody tasks with
different specificity and how introducing specific biological mechanisms to the
pre-training process can benefit the model. Additionally, we evaluate if the
learned antibody pre-trained representations can be applied to real-world
antibody problems, like drug discovery and immune process understanding.
Previously, no benchmark available largely hindered the study to answer these
questions. To aid in our investigation, we provide an AnTibody Understanding
Evaluation (ATUE) benchmark. We comprehensively evaluate the performance of
protein pre-trained language models by empirical study along with conclusions
and new insights. Our ATUE and code are released at
https://github.com/dqwang122/EATLM.
|
2302.04500 | Hexiang Pan | Hexiang Pan, Quang-Trung Ta, Meihui Zhang, Yeow Meng Chee, Gang Chen,
Beng Chin Ooi | FLAC: A Robust Failure-Aware Atomic Commit Protocol for Distributed
Transactions | null | null | null | null | cs.DC cs.AI cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In distributed transaction processing, atomic commit protocol (ACP) is used
to ensure database consistency. With the use of commodity compute nodes and
networks, failures such as system crashes and network partitioning are common.
It is therefore important for ACP to dynamically adapt to the operating
condition for efficiency while ensuring the consistency of the database.
Existing ACPs often assume stable operating conditions, hence, they are either
non-generalizable to different environments or slow in practice.
In this paper, we propose a novel and practical ACP, called Failure-Aware
Atomic Commit (FLAC). In essence, FLAC includes three protocols, which are
specifically designed for three different environments: (i) no failure occurs,
(ii) participant nodes might crash but there is no delayed connection, or (iii)
both crashed nodes and delayed connection can occur. It models these
environments as the failure-free, crash-failure, and network-failure robustness
levels. During its operation, FLAC can monitor if any failure occurs and
dynamically switch to operate the most suitable protocol, using a robustness
level state machine, whose parameters are fine-tuned by reinforcement learning.
Consequently, it improves both the response time and throughput, and
effectively handles nodes distributed across the Internet where crash and
network failures might occur. We implement FLAC in a distributed transactional
key-value storage system based on Google Percolator and evaluate its
performance with both a micro benchmark and a macro benchmark of real workload.
The results show that FLAC achieves up to 2.22x throughput improvement and
2.82x latency speedup, compared to existing ACPs for high-contention workloads.
| [
{
"version": "v1",
"created": "Thu, 9 Feb 2023 08:52:11 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Feb 2023 13:15:16 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 06:44:41 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Pan",
"Hexiang",
""
],
[
"Ta",
"Quang-Trung",
""
],
[
"Zhang",
"Meihui",
""
],
[
"Chee",
"Yeow Meng",
""
],
[
"Chen",
"Gang",
""
],
[
"Ooi",
"Beng Chin",
""
]
] | TITLE: FLAC: A Robust Failure-Aware Atomic Commit Protocol for Distributed
Transactions
ABSTRACT: In distributed transaction processing, atomic commit protocol (ACP) is used
to ensure database consistency. With the use of commodity compute nodes and
networks, failures such as system crashes and network partitioning are common.
It is therefore important for ACP to dynamically adapt to the operating
condition for efficiency while ensuring the consistency of the database.
Existing ACPs often assume stable operating conditions, hence, they are either
non-generalizable to different environments or slow in practice.
In this paper, we propose a novel and practical ACP, called Failure-Aware
Atomic Commit (FLAC). In essence, FLAC includes three protocols, which are
specifically designed for three different environments: (i) no failure occurs,
(ii) participant nodes might crash but there is no delayed connection, or (iii)
both crashed nodes and delayed connection can occur. It models these
environments as the failure-free, crash-failure, and network-failure robustness
levels. During its operation, FLAC can monitor if any failure occurs and
dynamically switch to operate the most suitable protocol, using a robustness
level state machine, whose parameters are fine-tuned by reinforcement learning.
Consequently, it improves both the response time and throughput, and
effectively handles nodes distributed across the Internet where crash and
network failures might occur. We implement FLAC in a distributed transactional
key-value storage system based on Google Percolator and evaluate its
performance with both a micro benchmark and a macro benchmark of real workload.
The results show that FLAC achieves up to 2.22x throughput improvement and
2.82x latency speedup, compared to existing ACPs for high-contention workloads.
|
2302.10198 | Qihuang Zhong | Qihuang Zhong, Liang Ding, Juhua Liu, Bo Du and Dacheng Tao | Can ChatGPT Understand Too? A Comparative Study on ChatGPT and
Fine-tuned BERT | Work in progress. Added results of advanced prompting strategies,
e.g., CoT. (19 pages) | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, ChatGPT has attracted great attention, as it can generate fluent
and high-quality responses to human inquiries. Several prior studies have shown
that ChatGPT attains remarkable generation ability compared with existing
models. However, the quantitative analysis of ChatGPT's understanding ability
has been given little attention. In this report, we explore the understanding
ability of ChatGPT by evaluating it on the most popular GLUE benchmark, and
comparing it with 4 representative fine-tuned BERT-style models. We find that:
1) ChatGPT falls short in handling paraphrase and similarity tasks; 2) ChatGPT
outperforms all BERT models on inference tasks by a large margin; 3) ChatGPT
achieves comparable performance compared with BERT on sentiment analysis and
question-answering tasks. Additionally, by combining some advanced prompting
strategies, we show that the understanding ability of ChatGPT can be further
improved.
| [
{
"version": "v1",
"created": "Sun, 19 Feb 2023 12:29:33 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 14:33:12 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Zhong",
"Qihuang",
""
],
[
"Ding",
"Liang",
""
],
[
"Liu",
"Juhua",
""
],
[
"Du",
"Bo",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Can ChatGPT Understand Too? A Comparative Study on ChatGPT and
Fine-tuned BERT
ABSTRACT: Recently, ChatGPT has attracted great attention, as it can generate fluent
and high-quality responses to human inquiries. Several prior studies have shown
that ChatGPT attains remarkable generation ability compared with existing
models. However, the quantitative analysis of ChatGPT's understanding ability
has been given little attention. In this report, we explore the understanding
ability of ChatGPT by evaluating it on the most popular GLUE benchmark, and
comparing it with 4 representative fine-tuned BERT-style models. We find that:
1) ChatGPT falls short in handling paraphrase and similarity tasks; 2) ChatGPT
outperforms all BERT models on inference tasks by a large margin; 3) ChatGPT
achieves comparable performance compared with BERT on sentiment analysis and
question-answering tasks. Additionally, by combining some advanced prompting
strategies, we show that the understanding ability of ChatGPT can be further
improved.
|
2302.10518 | Yuxuan Xiong | Yue Shi, Yuxuan Xiong, Jingyi Chai, Bingbing Ni, Wenjun Zhang | USR: Unsupervised Separated 3D Garment and Human Reconstruction via
Geometry and Semantic Consistency | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dressed people reconstruction from images is a popular task with promising
applications in the creative media and game industry. However, most existing
methods reconstruct the human body and garments as a whole with the supervision
of 3D models, which hinders the downstream interaction tasks and requires
hard-to-obtain data. To address these issues, we propose an unsupervised
separated 3D garments and human reconstruction model (USR), which reconstructs
the human body and authentic textured clothes in layers without 3D models. More
specifically, our method proposes a generalized surface-aware neural radiance
field to learn the mapping between sparse multi-view images and geometries of
the dressed people. Based on the full geometry, we introduce a Semantic and
Confidence Guided Separation strategy (SCGS) to detect, segment, and
reconstruct the clothes layer, leveraging the consistency between 2D semantic
and 3D geometry. Moreover, we propose a Geometry Fine-tune Module to smooth
edges. Extensive experiments on our dataset show that comparing with
state-of-the-art methods, USR achieves improvements on both geometry and
appearance reconstruction while supporting generalizing to unseen people in
real time. Besides, we also introduce SMPL-D model to show the benefit of the
separated modeling of clothes and the human body that allows swapping clothes
and virtual try-on.
| [
{
"version": "v1",
"created": "Tue, 21 Feb 2023 08:48:27 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Feb 2023 07:07:16 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 14:06:01 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shi",
"Yue",
""
],
[
"Xiong",
"Yuxuan",
""
],
[
"Chai",
"Jingyi",
""
],
[
"Ni",
"Bingbing",
""
],
[
"Zhang",
"Wenjun",
""
]
] | TITLE: USR: Unsupervised Separated 3D Garment and Human Reconstruction via
Geometry and Semantic Consistency
ABSTRACT: Dressed people reconstruction from images is a popular task with promising
applications in the creative media and game industry. However, most existing
methods reconstruct the human body and garments as a whole with the supervision
of 3D models, which hinders the downstream interaction tasks and requires
hard-to-obtain data. To address these issues, we propose an unsupervised
separated 3D garments and human reconstruction model (USR), which reconstructs
the human body and authentic textured clothes in layers without 3D models. More
specifically, our method proposes a generalized surface-aware neural radiance
field to learn the mapping between sparse multi-view images and geometries of
the dressed people. Based on the full geometry, we introduce a Semantic and
Confidence Guided Separation strategy (SCGS) to detect, segment, and
reconstruct the clothes layer, leveraging the consistency between 2D semantic
and 3D geometry. Moreover, we propose a Geometry Fine-tune Module to smooth
edges. Extensive experiments on our dataset show that comparing with
state-of-the-art methods, USR achieves improvements on both geometry and
appearance reconstruction while supporting generalizing to unseen people in
real time. Besides, we also introduce SMPL-D model to show the benefit of the
separated modeling of clothes and the human body that allows swapping clothes
and virtual try-on.
|
2302.11077 | Fred Song | Yu Song, Madhav V. Chitturi, David A. Noyce | Impact of Event Encoding and Dissimilarity Measures on Traffic Crash
Characterization Based on Sequence of Events | null | null | 10.1016/j.aap.2023.107016 | null | stat.AP cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Crash sequence analysis has been shown in prior studies to be useful for
characterizing crashes and identifying safety countermeasures. Sequence
analysis is highly domain-specific, but its various techniques have not been
evaluated for adaptation to crash sequences. This paper evaluates the impact of
encoding and dissimilarity measures on crash sequence analysis and clustering.
Sequence data of interstate highway, single-vehicle crashes in the United
States, from 2016-2018, were studied. Two encoding schemes and five optimal
matching based dissimilarity measures were compared by evaluating the sequence
clustering results. The five dissimilarity measures were categorized into two
groups based on correlations between dissimilarity matrices. The optimal
dissimilarity measure and encoding scheme were identified based on the
agreements with a benchmark crash categorization. The transition-rate-based,
localized optimal matching dissimilarity and consolidated encoding scheme had
the highest agreement with the benchmark. Evaluation results indicate that the
selection of dissimilarity measure and encoding scheme determines the results
of sequence clustering and crash characterization. A dissimilarity measure that
considers the relationships between events and domain context tends to perform
well in crash sequence clustering. An encoding scheme that consolidates similar
events naturally takes domain context into consideration.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 00:38:15 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Song",
"Yu",
""
],
[
"Chitturi",
"Madhav V.",
""
],
[
"Noyce",
"David A.",
""
]
] | TITLE: Impact of Event Encoding and Dissimilarity Measures on Traffic Crash
Characterization Based on Sequence of Events
ABSTRACT: Crash sequence analysis has been shown in prior studies to be useful for
characterizing crashes and identifying safety countermeasures. Sequence
analysis is highly domain-specific, but its various techniques have not been
evaluated for adaptation to crash sequences. This paper evaluates the impact of
encoding and dissimilarity measures on crash sequence analysis and clustering.
Sequence data of interstate highway, single-vehicle crashes in the United
States, from 2016-2018, were studied. Two encoding schemes and five optimal
matching based dissimilarity measures were compared by evaluating the sequence
clustering results. The five dissimilarity measures were categorized into two
groups based on correlations between dissimilarity matrices. The optimal
dissimilarity measure and encoding scheme were identified based on the
agreements with a benchmark crash categorization. The transition-rate-based,
localized optimal matching dissimilarity and consolidated encoding scheme had
the highest agreement with the benchmark. Evaluation results indicate that the
selection of dissimilarity measure and encoding scheme determines the results
of sequence clustering and crash characterization. A dissimilarity measure that
considers the relationships between events and domain context tends to perform
well in crash sequence clustering. An encoding scheme that consolidates similar
events naturally takes domain context into consideration.
|
2302.14557 | Axi Niu | Axi Niu, Pei Wang, Yu Zhu, Jinqiu Sun, Qingsen Yan, Yanning Zhang | GRAN: Ghost Residual Attention Network for Single Image Super Resolution | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recently, many works have designed wider and deeper networks to achieve
higher image super-resolution performance. Despite their outstanding
performance, they still suffer from high computational resources, preventing
them from directly applying to embedded devices. To reduce the computation
resources and maintain performance, we propose a novel Ghost Residual Attention
Network (GRAN) for efficient super-resolution. This paper introduces Ghost
Residual Attention Block (GRAB) groups to overcome the drawbacks of the
standard convolutional operation, i.e., redundancy of the intermediate feature.
GRAB consists of the Ghost Module and Channel and Spatial Attention Module
(CSAM) to alleviate the generation of redundant features. Specifically, Ghost
Module can reveal information underlying intrinsic features by employing linear
operations to replace the standard convolutions. Reducing redundant features by
the Ghost Module, our model decreases memory and computing resource
requirements in the network. The CSAM pays more comprehensive attention to
where and what the feature extraction is, which is critical to recovering the
image details. Experiments conducted on the benchmark datasets demonstrate the
superior performance of our method in both qualitative and quantitative.
Compared to the baseline models, we achieve higher performance with lower
computational resources, whose parameters and FLOPs have decreased by more than
ten times.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2023 13:26:24 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 02:01:19 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Niu",
"Axi",
""
],
[
"Wang",
"Pei",
""
],
[
"Zhu",
"Yu",
""
],
[
"Sun",
"Jinqiu",
""
],
[
"Yan",
"Qingsen",
""
],
[
"Zhang",
"Yanning",
""
]
] | TITLE: GRAN: Ghost Residual Attention Network for Single Image Super Resolution
ABSTRACT: Recently, many works have designed wider and deeper networks to achieve
higher image super-resolution performance. Despite their outstanding
performance, they still suffer from high computational resources, preventing
them from directly applying to embedded devices. To reduce the computation
resources and maintain performance, we propose a novel Ghost Residual Attention
Network (GRAN) for efficient super-resolution. This paper introduces Ghost
Residual Attention Block (GRAB) groups to overcome the drawbacks of the
standard convolutional operation, i.e., redundancy of the intermediate feature.
GRAB consists of the Ghost Module and Channel and Spatial Attention Module
(CSAM) to alleviate the generation of redundant features. Specifically, Ghost
Module can reveal information underlying intrinsic features by employing linear
operations to replace the standard convolutions. Reducing redundant features by
the Ghost Module, our model decreases memory and computing resource
requirements in the network. The CSAM pays more comprehensive attention to
where and what the feature extraction is, which is critical to recovering the
image details. Experiments conducted on the benchmark datasets demonstrate the
superior performance of our method in both qualitative and quantitative.
Compared to the baseline models, we achieve higher performance with lower
computational resources, whose parameters and FLOPs have decreased by more than
ten times.
|
2303.00175 | Dimitrios Kollias | Dimitrios Kollias and Anastasios Arsenos and Stefanos Kollias | A Deep Neural Architecture for Harmonizing 3-D Input Data Analysis and
Decision Making in Medical Imaging | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Harmonizing the analysis of data, especially of 3-D image volumes, consisting
of different number of slices and annotated per volume, is a significant
problem in training and using deep neural networks in various applications,
including medical imaging. Moreover, unifying the decision making of the
networks over different input datasets is crucial for the generation of rich
data-driven knowledge and for trusted usage in the applications. This paper
presents a new deep neural architecture, named RACNet, which includes routing
and feature alignment steps and effectively handles different input lengths and
single annotations of the 3-D image inputs, whilst providing highly accurate
decisions. In addition, through latent variable extraction from the trained
RACNet, a set of anchors are generated providing further insight on the
network's decision making. These can be used to enrich and unify data-driven
knowledge extracted from different datasets. An extensive experimental study
illustrates the above developments, focusing on COVID-19 diagnosis through
analysis of 3-D chest CT scans from databases generated in different countries
and medical centers.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 02:07:48 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 01:40:50 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kollias",
"Dimitrios",
""
],
[
"Arsenos",
"Anastasios",
""
],
[
"Kollias",
"Stefanos",
""
]
] | TITLE: A Deep Neural Architecture for Harmonizing 3-D Input Data Analysis and
Decision Making in Medical Imaging
ABSTRACT: Harmonizing the analysis of data, especially of 3-D image volumes, consisting
of different number of slices and annotated per volume, is a significant
problem in training and using deep neural networks in various applications,
including medical imaging. Moreover, unifying the decision making of the
networks over different input datasets is crucial for the generation of rich
data-driven knowledge and for trusted usage in the applications. This paper
presents a new deep neural architecture, named RACNet, which includes routing
and feature alignment steps and effectively handles different input lengths and
single annotations of the 3-D image inputs, whilst providing highly accurate
decisions. In addition, through latent variable extraction from the trained
RACNet, a set of anchors are generated providing further insight on the
network's decision making. These can be used to enrich and unify data-driven
knowledge extracted from different datasets. An extensive experimental study
illustrates the above developments, focusing on COVID-19 diagnosis through
analysis of 3-D chest CT scans from databases generated in different countries
and medical centers.
|
2303.00199 | Kun Yang | Kun Yang, Jun Lu | DMSA: Dynamic Multi-scale Unsupervised Semantic Segmentation Based on
Adaptive Affinity | 5 pages,4 figures | ICASSP 2023 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The proposed method in this paper proposes an end-to-end unsupervised
semantic segmentation architecture DMSA based on four loss functions. The
framework uses Atrous Spatial Pyramid Pooling (ASPP) module to enhance feature
extraction. At the same time, a dynamic dilation strategy is designed to better
capture multi-scale context information. Secondly, a Pixel-Adaptive Refinement
(PAR) module is introduced, which can adaptively refine the initial pseudo
labels after feature fusion to obtain high quality pseudo labels. Experiments
show that the proposed DSMA framework is superior to the existing methods on
the saliency dataset. On the COCO 80 dataset, the MIoU is improved by 2.0, and
the accuracy is improved by 5.39. On the Pascal VOC 2012 Augmented dataset, the
MIoU is improved by 4.9, and the accuracy is improved by 3.4. In addition, the
convergence speed of the model is also greatly improved after the introduction
of the PAR module.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 03:08:30 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yang",
"Kun",
""
],
[
"Lu",
"Jun",
""
]
] | TITLE: DMSA: Dynamic Multi-scale Unsupervised Semantic Segmentation Based on
Adaptive Affinity
ABSTRACT: The proposed method in this paper proposes an end-to-end unsupervised
semantic segmentation architecture DMSA based on four loss functions. The
framework uses Atrous Spatial Pyramid Pooling (ASPP) module to enhance feature
extraction. At the same time, a dynamic dilation strategy is designed to better
capture multi-scale context information. Secondly, a Pixel-Adaptive Refinement
(PAR) module is introduced, which can adaptively refine the initial pseudo
labels after feature fusion to obtain high quality pseudo labels. Experiments
show that the proposed DSMA framework is superior to the existing methods on
the saliency dataset. On the COCO 80 dataset, the MIoU is improved by 2.0, and
the accuracy is improved by 5.39. On the Pascal VOC 2012 Augmented dataset, the
MIoU is improved by 4.9, and the accuracy is improved by 3.4. In addition, the
convergence speed of the model is also greatly improved after the introduction
of the PAR module.
|
2303.00216 | Naoki Akai | Naoki Akai | Efficient Solution to 3D-LiDAR-based Monte Carlo Localization with
Fusion of Measurement Model Optimization via Importance Sampling | 8 pages, 4 figures | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This paper presents an efficient solution to 3D-LiDAR-based Monte Carlo
localization (MCL). MCL robustly works if particles are exactly sampled around
the ground truth. An inertial navigation system (INS) can be used for accurate
sampling, but many particles are still needed to be used for solving the 3D
localization problem even if INS is available. In particular, huge number of
particles are necessary if INS is not available and it makes infeasible to
perform 3D MCL in terms of the computational cost. Scan matching (SM), that is
optimization-based localization, efficiently works even though INS is not
available because SM can ignore movement constraints of a robot and/or device
in its optimization process. However, SM sometimes determines an infeasible
estimate against movement. We consider that MCL and SM have complemental
advantages and disadvantages and propose a fusion method of MCL and SM. Because
SM is considered as optimization of a measurement model in terms of the
probabilistic modeling, we perform measurement model optimization as SM. The
optimization result is then used to approximate the measurement model
distribution and the approximated distribution is used to sample particles. The
sampled particles are fused with MCL via importance sampling. As a result, the
advantages of MCL and SM can be simultaneously utilized while mitigating their
disadvantages. Experiments are conducted on the KITTI dataset and other two
open datasets. Results show that the presented method can be run on a single
CPU thread and accurately perform localization even if INS is not available.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 03:40:31 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 06:31:13 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Akai",
"Naoki",
""
]
] | TITLE: Efficient Solution to 3D-LiDAR-based Monte Carlo Localization with
Fusion of Measurement Model Optimization via Importance Sampling
ABSTRACT: This paper presents an efficient solution to 3D-LiDAR-based Monte Carlo
localization (MCL). MCL robustly works if particles are exactly sampled around
the ground truth. An inertial navigation system (INS) can be used for accurate
sampling, but many particles are still needed to be used for solving the 3D
localization problem even if INS is available. In particular, huge number of
particles are necessary if INS is not available and it makes infeasible to
perform 3D MCL in terms of the computational cost. Scan matching (SM), that is
optimization-based localization, efficiently works even though INS is not
available because SM can ignore movement constraints of a robot and/or device
in its optimization process. However, SM sometimes determines an infeasible
estimate against movement. We consider that MCL and SM have complemental
advantages and disadvantages and propose a fusion method of MCL and SM. Because
SM is considered as optimization of a measurement model in terms of the
probabilistic modeling, we perform measurement model optimization as SM. The
optimization result is then used to approximate the measurement model
distribution and the approximated distribution is used to sample particles. The
sampled particles are fused with MCL via importance sampling. As a result, the
advantages of MCL and SM can be simultaneously utilized while mitigating their
disadvantages. Experiments are conducted on the KITTI dataset and other two
open datasets. Results show that the presented method can be run on a single
CPU thread and accurately perform localization even if INS is not available.
|
2303.00232 | Xiaoyu Liu | Xiaoyu Liu, Linhao Qu, Ziyue Xie, Jiayue Zhao, Yonghong Shi, and
Zhijian Song | Towards more precise automatic analysis: a comprehensive survey of deep
learning-based multi-organ segmentation | 25 pages, 9 figures, 16 tabels | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate segmentation of multiple organs of the head, neck, chest, and
abdomen from medical images is an essential step in computer-aided diagnosis,
surgical navigation, and radiation therapy. In the past few years, with a
data-driven feature extraction approach and end-to-end training, automatic deep
learning-based multi-organ segmentation method has far outperformed traditional
methods and become a new research topic. This review systematically summarizes
the latest research in this field. For the first time, from the perspective of
full and imperfect annotation, we comprehensively compile 161 studies on deep
learning-based multi-organ segmentation in multiple regions such as the head
and neck, chest, and abdomen, containing a total of 214 related references. The
method based on full annotation summarizes the existing methods from four
aspects: network architecture, network dimension, network dedicated modules,
and network loss function. The method based on imperfect annotation summarizes
the existing methods from two aspects: weak annotation-based methods and semi
annotation-based methods. We also summarize frequently used datasets for
multi-organ segmentation and discuss new challenges and new research trends in
this field.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 04:52:49 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 10:02:03 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Liu",
"Xiaoyu",
""
],
[
"Qu",
"Linhao",
""
],
[
"Xie",
"Ziyue",
""
],
[
"Zhao",
"Jiayue",
""
],
[
"Shi",
"Yonghong",
""
],
[
"Song",
"Zhijian",
""
]
] | TITLE: Towards more precise automatic analysis: a comprehensive survey of deep
learning-based multi-organ segmentation
ABSTRACT: Accurate segmentation of multiple organs of the head, neck, chest, and
abdomen from medical images is an essential step in computer-aided diagnosis,
surgical navigation, and radiation therapy. In the past few years, with a
data-driven feature extraction approach and end-to-end training, automatic deep
learning-based multi-organ segmentation method has far outperformed traditional
methods and become a new research topic. This review systematically summarizes
the latest research in this field. For the first time, from the perspective of
full and imperfect annotation, we comprehensively compile 161 studies on deep
learning-based multi-organ segmentation in multiple regions such as the head
and neck, chest, and abdomen, containing a total of 214 related references. The
method based on full annotation summarizes the existing methods from four
aspects: network architecture, network dimension, network dedicated modules,
and network loss function. The method based on imperfect annotation summarizes
the existing methods from two aspects: weak annotation-based methods and semi
annotation-based methods. We also summarize frequently used datasets for
multi-organ segmentation and discuss new challenges and new research trends in
this field.
|
2303.00369 | Zhang Jingyi | Lingke Kong, X. Sharon Qi, Qijin Shen, Jiacheng Wang, Jingyi Zhang,
Yanle Hu, Qichao Zhou | Indescribable Multi-modal Spatial Evaluator | Accepted by CVPR2023 | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Multi-modal image registration spatially aligns two images with different
distributions. One of its major challenges is that images acquired from
different imaging machines have different imaging distributions, making it
difficult to focus only on the spatial aspect of the images and ignore
differences in distributions. In this study, we developed a self-supervised
approach, Indescribable Multi-model Spatial Evaluator (IMSE), to address
multi-modal image registration. IMSE creates an accurate multi-modal spatial
evaluator to measure spatial differences between two images, and then optimizes
registration by minimizing the error predicted of the evaluator. To optimize
IMSE performance, we also proposed a new style enhancement method called
Shuffle Remap which randomizes the image distribution into multiple segments,
and then randomly disorders and remaps these segments, so that the distribution
of the original image is changed. Shuffle Remap can help IMSE to predict the
difference in spatial location from unseen target distributions. Our results
show that IMSE outperformed the existing methods for registration using T1-T2
and CT-MRI datasets. IMSE also can be easily integrated into the traditional
registration process, and can provide a convenient way to evaluate and
visualize registration results. IMSE also has the potential to be used as a new
paradigm for image-to-image translation. Our code is available at
https://github.com/Kid-Liet/IMSE.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 09:50:39 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 01:23:23 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kong",
"Lingke",
""
],
[
"Qi",
"X. Sharon",
""
],
[
"Shen",
"Qijin",
""
],
[
"Wang",
"Jiacheng",
""
],
[
"Zhang",
"Jingyi",
""
],
[
"Hu",
"Yanle",
""
],
[
"Zhou",
"Qichao",
""
]
] | TITLE: Indescribable Multi-modal Spatial Evaluator
ABSTRACT: Multi-modal image registration spatially aligns two images with different
distributions. One of its major challenges is that images acquired from
different imaging machines have different imaging distributions, making it
difficult to focus only on the spatial aspect of the images and ignore
differences in distributions. In this study, we developed a self-supervised
approach, Indescribable Multi-model Spatial Evaluator (IMSE), to address
multi-modal image registration. IMSE creates an accurate multi-modal spatial
evaluator to measure spatial differences between two images, and then optimizes
registration by minimizing the error predicted of the evaluator. To optimize
IMSE performance, we also proposed a new style enhancement method called
Shuffle Remap which randomizes the image distribution into multiple segments,
and then randomly disorders and remaps these segments, so that the distribution
of the original image is changed. Shuffle Remap can help IMSE to predict the
difference in spatial location from unseen target distributions. Our results
show that IMSE outperformed the existing methods for registration using T1-T2
and CT-MRI datasets. IMSE also can be easily integrated into the traditional
registration process, and can provide a convenient way to evaluate and
visualize registration results. IMSE also has the potential to be used as a new
paradigm for image-to-image translation. Our code is available at
https://github.com/Kid-Liet/IMSE.
|
2303.00409 | Ming-Chang Lee | Ming-Chang Lee and Jia-Chun Lin | RePAD2: Real-Time, Lightweight, and Adaptive Anomaly Detection for
Open-Ended Time Series | 10 pages, 11 figures, and 10 tables, the paper is accepted by 8th
International Conference on Internet of Things, Big Data and Security (IoTBDS
2023) | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An open-ended time series refers to a series of data points indexed in time
order without an end. Such a time series can be found everywhere due to the
prevalence of Internet of Things. Providing lightweight and real-time anomaly
detection for open-ended time series is highly desirable to industry and
organizations since it allows immediate response and avoids potential financial
loss. In the last few years, several real-time time series anomaly detection
approaches have been introduced. However, they might exhaust system resources
when they are applied to open-ended time series for a long time. To address
this issue, in this paper we propose RePAD2, a lightweight real-time anomaly
detection approach for open-ended time series by improving its predecessor
RePAD, which is one of the state-of-the-art anomaly detection approaches. We
conducted a series of experiments to compare RePAD2 with RePAD and another
similar detection approach based on real-world time series datasets, and
demonstrated that RePAD2 can address the mentioned resource exhaustion issue
while offering comparable detection accuracy and slightly less time
consumption.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 11:00:20 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 08:04:03 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lee",
"Ming-Chang",
""
],
[
"Lin",
"Jia-Chun",
""
]
] | TITLE: RePAD2: Real-Time, Lightweight, and Adaptive Anomaly Detection for
Open-Ended Time Series
ABSTRACT: An open-ended time series refers to a series of data points indexed in time
order without an end. Such a time series can be found everywhere due to the
prevalence of Internet of Things. Providing lightweight and real-time anomaly
detection for open-ended time series is highly desirable to industry and
organizations since it allows immediate response and avoids potential financial
loss. In the last few years, several real-time time series anomaly detection
approaches have been introduced. However, they might exhaust system resources
when they are applied to open-ended time series for a long time. To address
this issue, in this paper we propose RePAD2, a lightweight real-time anomaly
detection approach for open-ended time series by improving its predecessor
RePAD, which is one of the state-of-the-art anomaly detection approaches. We
conducted a series of experiments to compare RePAD2 with RePAD and another
similar detection approach based on real-world time series datasets, and
demonstrated that RePAD2 can address the mentioned resource exhaustion issue
while offering comparable detection accuracy and slightly less time
consumption.
|
2303.00563 | Depu Meng | Depu Meng, Owen Sayer, Rusheng Zhang, Shengyin Shen, Houqiang Li,
Henry X. Liu | ROCO: A Roundabout Traffic Conflict Dataset | Accepted by TRBAM 2023 presentation | null | null | null | eess.SY cs.CV cs.RO cs.SY eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traffic conflicts have been studied by the transportation research community
as a surrogate safety measure for decades. However, due to the rarity of
traffic conflicts, collecting large-scale real-world traffic conflict data
becomes extremely challenging. In this paper, we introduce and analyze ROCO - a
real-world roundabout traffic conflict dataset. The data is collected at a
two-lane roundabout at the intersection of State St. and W. Ellsworth Rd. in
Ann Arbor, Michigan. We use raw video dataflow captured from four fisheye
cameras installed at the roundabout as our input data source. We adopt a
learning-based conflict identification algorithm from video to find potential
traffic conflicts, and then manually label them for dataset collection and
annotation. In total 557 traffic conflicts and 17 traffic crashes are collected
from August 2021 to October 2021. We provide trajectory data of the traffic
conflict scenes extracted using our roadside perception system. Taxonomy based
on traffic conflict severity, reason for the traffic conflict, and its effect
on the traffic flow is provided. With the traffic conflict data collected, we
discover that failure to yield to circulating vehicles when entering the
roundabout is the largest contributing reason for traffic conflicts. ROCO
dataset will be made public in the short future.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 15:09:45 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 02:26:07 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Meng",
"Depu",
""
],
[
"Sayer",
"Owen",
""
],
[
"Zhang",
"Rusheng",
""
],
[
"Shen",
"Shengyin",
""
],
[
"Li",
"Houqiang",
""
],
[
"Liu",
"Henry X.",
""
]
] | TITLE: ROCO: A Roundabout Traffic Conflict Dataset
ABSTRACT: Traffic conflicts have been studied by the transportation research community
as a surrogate safety measure for decades. However, due to the rarity of
traffic conflicts, collecting large-scale real-world traffic conflict data
becomes extremely challenging. In this paper, we introduce and analyze ROCO - a
real-world roundabout traffic conflict dataset. The data is collected at a
two-lane roundabout at the intersection of State St. and W. Ellsworth Rd. in
Ann Arbor, Michigan. We use raw video dataflow captured from four fisheye
cameras installed at the roundabout as our input data source. We adopt a
learning-based conflict identification algorithm from video to find potential
traffic conflicts, and then manually label them for dataset collection and
annotation. In total 557 traffic conflicts and 17 traffic crashes are collected
from August 2021 to October 2021. We provide trajectory data of the traffic
conflict scenes extracted using our roadside perception system. Taxonomy based
on traffic conflict severity, reason for the traffic conflict, and its effect
on the traffic flow is provided. With the traffic conflict data collected, we
discover that failure to yield to circulating vehicles when entering the
roundabout is the largest contributing reason for traffic conflicts. ROCO
dataset will be made public in the short future.
|
2303.00755 | Christian Marius Lillelund | Christian Marius Lillelund, Henrik Bagger Jensen, Christian Fischer
Pedersen | Cloud K-SVD for Image Denoising | null | SN COMPUT. SCI. 3, 151 (2022) | 10.1007/s42979-022-01042-y | null | eess.IV cs.AI cs.CV cs.DC cs.LG | http://creativecommons.org/licenses/by/4.0/ | Cloud K-SVD is a dictionary learning algorithm that can train at multiple
nodes and hereby produce a mutual dictionary to represent low-dimensional
geometric structures in image data. We present a novel application of the
algorithm as we use it to recover both noiseless and noisy images from
overlapping patches. We implement a node network in Kubernetes using Docker
containers to facilitate Cloud K-SVD. Results show that Cloud K-SVD can recover
images approximately and remove quantifiable amounts of noise from benchmark
gray-scaled images without sacrificing accuracy in recovery; we achieve an SSIM
index of 0.88, 0.91 and 0.95 between clean and recovered images for noise
levels ($\mu$ = 0, $\sigma^{2}$ = 0.01, 0.005, 0.001), respectively, which is
similar to SOTA in the field. Cloud K-SVD is evidently able to learn a mutual
dictionary across multiple nodes and remove AWGN from images. The mutual
dictionary can be used to recover a specific image at any of the nodes in the
network.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 15:44:55 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Lillelund",
"Christian Marius",
""
],
[
"Jensen",
"Henrik Bagger",
""
],
[
"Pedersen",
"Christian Fischer",
""
]
] | TITLE: Cloud K-SVD for Image Denoising
ABSTRACT: Cloud K-SVD is a dictionary learning algorithm that can train at multiple
nodes and hereby produce a mutual dictionary to represent low-dimensional
geometric structures in image data. We present a novel application of the
algorithm as we use it to recover both noiseless and noisy images from
overlapping patches. We implement a node network in Kubernetes using Docker
containers to facilitate Cloud K-SVD. Results show that Cloud K-SVD can recover
images approximately and remove quantifiable amounts of noise from benchmark
gray-scaled images without sacrificing accuracy in recovery; we achieve an SSIM
index of 0.88, 0.91 and 0.95 between clean and recovered images for noise
levels ($\mu$ = 0, $\sigma^{2}$ = 0.01, 0.005, 0.001), respectively, which is
similar to SOTA in the field. Cloud K-SVD is evidently able to learn a mutual
dictionary across multiple nodes and remove AWGN from images. The mutual
dictionary can be used to recover a specific image at any of the nodes in the
network.
|
2303.00802 | Philipp Klumpp | Philipp Klumpp, Pooja Chitkara, Leda Sar{\i}, Prashant Serai, Jilong
Wu, Irina-Elena Veliche, Rongqing Huang, Qing He | Synthetic Cross-accent Data Augmentation for Automatic Speech
Recognition | null | null | null | null | cs.CL cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The awareness for biased ASR datasets or models has increased notably in
recent years. Even for English, despite a vast amount of available training
data, systems perform worse for non-native speakers. In this work, we improve
an accent-conversion model (ACM) which transforms native US-English speech into
accented pronunciation. We include phonetic knowledge in the ACM training to
provide accurate feedback about how well certain pronunciation patterns were
recovered in the synthesized waveform. Furthermore, we investigate the
feasibility of learned accent representations instead of static embeddings.
Generated data was then used to train two state-of-the-art ASR systems. We
evaluated our approach on native and non-native English datasets and found that
synthetically accented data helped the ASR to better understand speech from
seen accents. This observation did not translate to unseen accents, and it was
not observed for a model that had been pre-trained exclusively with native
speech.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 20:05:19 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Klumpp",
"Philipp",
""
],
[
"Chitkara",
"Pooja",
""
],
[
"Sarı",
"Leda",
""
],
[
"Serai",
"Prashant",
""
],
[
"Wu",
"Jilong",
""
],
[
"Veliche",
"Irina-Elena",
""
],
[
"Huang",
"Rongqing",
""
],
[
"He",
"Qing",
""
]
] | TITLE: Synthetic Cross-accent Data Augmentation for Automatic Speech
Recognition
ABSTRACT: The awareness for biased ASR datasets or models has increased notably in
recent years. Even for English, despite a vast amount of available training
data, systems perform worse for non-native speakers. In this work, we improve
an accent-conversion model (ACM) which transforms native US-English speech into
accented pronunciation. We include phonetic knowledge in the ACM training to
provide accurate feedback about how well certain pronunciation patterns were
recovered in the synthesized waveform. Furthermore, we investigate the
feasibility of learned accent representations instead of static embeddings.
Generated data was then used to train two state-of-the-art ASR systems. We
evaluated our approach on native and non-native English datasets and found that
synthetically accented data helped the ASR to better understand speech from
seen accents. This observation did not translate to unseen accents, and it was
not observed for a model that had been pre-trained exclusively with native
speech.
|
2303.00815 | Jingli Shi | Jingli Shi, Weihua Li, Quan Bai, Yi Yang, Jianhua Jiang | Soft Prompt Guided Joint Learning for Cross-Domain Sentiment Analysis | 22 pages | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Aspect term extraction is a fundamental task in fine-grained sentiment
analysis, which aims at detecting customer's opinion targets from reviews on
product or service. The traditional supervised models can achieve promising
results with annotated datasets, however, the performance dramatically
decreases when they are applied to the task of cross-domain aspect term
extraction. Existing cross-domain transfer learning methods either directly
inject linguistic features into Language models, making it difficult to
transfer linguistic knowledge to target domain, or rely on the fixed predefined
prompts, which is time-consuming to construct the prompts over all potential
aspect term spans. To resolve the limitations, we propose a soft prompt-based
joint learning method for cross domain aspect term extraction in this paper.
Specifically, by incorporating external linguistic features, the proposed
method learn domain-invariant representations between source and target domains
via multiple objectives, which bridges the gap between domains with varied
distributions of aspect terms. Further, the proposed method interpolates a set
of transferable soft prompts consisted of multiple learnable vectors that are
beneficial to detect aspect terms in target domain. Extensive experiments are
conducted on the benchmark datasets and the experimental results demonstrate
the effectiveness of the proposed method for cross-domain aspect terms
extraction.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 20:33:37 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shi",
"Jingli",
""
],
[
"Li",
"Weihua",
""
],
[
"Bai",
"Quan",
""
],
[
"Yang",
"Yi",
""
],
[
"Jiang",
"Jianhua",
""
]
] | TITLE: Soft Prompt Guided Joint Learning for Cross-Domain Sentiment Analysis
ABSTRACT: Aspect term extraction is a fundamental task in fine-grained sentiment
analysis, which aims at detecting customer's opinion targets from reviews on
product or service. The traditional supervised models can achieve promising
results with annotated datasets, however, the performance dramatically
decreases when they are applied to the task of cross-domain aspect term
extraction. Existing cross-domain transfer learning methods either directly
inject linguistic features into Language models, making it difficult to
transfer linguistic knowledge to target domain, or rely on the fixed predefined
prompts, which is time-consuming to construct the prompts over all potential
aspect term spans. To resolve the limitations, we propose a soft prompt-based
joint learning method for cross domain aspect term extraction in this paper.
Specifically, by incorporating external linguistic features, the proposed
method learn domain-invariant representations between source and target domains
via multiple objectives, which bridges the gap between domains with varied
distributions of aspect terms. Further, the proposed method interpolates a set
of transferable soft prompts consisted of multiple learnable vectors that are
beneficial to detect aspect terms in target domain. Extensive experiments are
conducted on the benchmark datasets and the experimental results demonstrate
the effectiveness of the proposed method for cross-domain aspect terms
extraction.
|
2303.00821 | Saksham Bassi | Aayush Agarwal and Saksham Bassi | Learning high-dimensional causal effect | null | null | null | null | cs.LG stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The scarcity of high-dimensional causal inference datasets restricts the
exploration of complex deep models. In this work, we propose a method to
generate a synthetic causal dataset that is high-dimensional. The synthetic
data simulates a causal effect using the MNIST dataset with Bernoulli treatment
values. This provides an opportunity to study varieties of models for causal
effect estimation. We experiment on this dataset using Dragonnet architecture
(Shi et al. (2019)) and modified architectures. We use the modified
architectures to explore different types of initial Neural Network layers and
observe that the modified architectures perform better in estimations. We
observe that residual and transformer models estimate treatment effect very
closely without the need for targeted regularization, introduced by Shi et al.
(2019).
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 20:57:48 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Agarwal",
"Aayush",
""
],
[
"Bassi",
"Saksham",
""
]
] | TITLE: Learning high-dimensional causal effect
ABSTRACT: The scarcity of high-dimensional causal inference datasets restricts the
exploration of complex deep models. In this work, we propose a method to
generate a synthetic causal dataset that is high-dimensional. The synthetic
data simulates a causal effect using the MNIST dataset with Bernoulli treatment
values. This provides an opportunity to study varieties of models for causal
effect estimation. We experiment on this dataset using Dragonnet architecture
(Shi et al. (2019)) and modified architectures. We use the modified
architectures to explore different types of initial Neural Network layers and
observe that the modified architectures perform better in estimations. We
observe that residual and transformer models estimate treatment effect very
closely without the need for targeted regularization, introduced by Shi et al.
(2019).
|
2303.00885 | Siyuan Yan | Siyuan Yan, Zhen Yu, Xuelin Zhang, Dwarikanath Mahapatra, Shekhar S.
Chandra, Monika Janda, Peter Soyer, Zongyuan Ge | Towards Trustable Skin Cancer Diagnosis via Rewriting Model's Decision | Accepted by CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep neural networks have demonstrated promising performance on image
recognition tasks. However, they may heavily rely on confounding factors, using
irrelevant artifacts or bias within the dataset as the cue to improve
performance. When a model performs decision-making based on these spurious
correlations, it can become untrustable and lead to catastrophic outcomes when
deployed in the real-world scene. In this paper, we explore and try to solve
this problem in the context of skin cancer diagnosis. We introduce a
human-in-the-loop framework in the model training process such that users can
observe and correct the model's decision logic when confounding behaviors
happen. Specifically, our method can automatically discover confounding factors
by analyzing the co-occurrence behavior of the samples. It is capable of
learning confounding concepts using easily obtained concept exemplars. By
mapping the black-box model's feature representation onto an explainable
concept space, human users can interpret the concept and intervene via first
order-logic instruction. We systematically evaluate our method on our newly
crafted, well-controlled skin lesion dataset and several public skin lesion
datasets. Experiments show that our method can effectively detect and remove
confounding factors from datasets without any prior knowledge about the
category distribution and does not require fully annotated concept labels. We
also show that our method enables the model to focus on clinical-related
concepts, improving the model's performance and trustworthiness during model
inference.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 01:02:18 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yan",
"Siyuan",
""
],
[
"Yu",
"Zhen",
""
],
[
"Zhang",
"Xuelin",
""
],
[
"Mahapatra",
"Dwarikanath",
""
],
[
"Chandra",
"Shekhar S.",
""
],
[
"Janda",
"Monika",
""
],
[
"Soyer",
"Peter",
""
],
[
"Ge",
"Zongyuan",
""
]
] | TITLE: Towards Trustable Skin Cancer Diagnosis via Rewriting Model's Decision
ABSTRACT: Deep neural networks have demonstrated promising performance on image
recognition tasks. However, they may heavily rely on confounding factors, using
irrelevant artifacts or bias within the dataset as the cue to improve
performance. When a model performs decision-making based on these spurious
correlations, it can become untrustable and lead to catastrophic outcomes when
deployed in the real-world scene. In this paper, we explore and try to solve
this problem in the context of skin cancer diagnosis. We introduce a
human-in-the-loop framework in the model training process such that users can
observe and correct the model's decision logic when confounding behaviors
happen. Specifically, our method can automatically discover confounding factors
by analyzing the co-occurrence behavior of the samples. It is capable of
learning confounding concepts using easily obtained concept exemplars. By
mapping the black-box model's feature representation onto an explainable
concept space, human users can interpret the concept and intervene via first
order-logic instruction. We systematically evaluate our method on our newly
crafted, well-controlled skin lesion dataset and several public skin lesion
datasets. Experiments show that our method can effectively detect and remove
confounding factors from datasets without any prior knowledge about the
category distribution and does not require fully annotated concept labels. We
also show that our method enables the model to focus on clinical-related
concepts, improving the model's performance and trustworthiness during model
inference.
|
2303.00897 | Dun Zeng | Dun Zeng, Xiangjing Hu, Shiyu Liu, Yue Yu, Qifan Wang, Zenglin Xu | Stochastic Clustered Federated Learning | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Federated learning is a distributed learning framework that takes full
advantage of private data samples kept on edge devices. In real-world federated
learning systems, these data samples are often decentralized and
Non-Independently Identically Distributed (Non-IID), causing divergence and
performance degradation in the federated learning process. As a new solution,
clustered federated learning groups federated clients with similar data
distributions to impair the Non-IID effects and train a better model for every
cluster. This paper proposes StoCFL, a novel clustered federated learning
approach for generic Non-IID issues. In detail, StoCFL implements a flexible
CFL framework that supports an arbitrary proportion of client participation and
newly joined clients for a varying FL system, while maintaining a great
improvement in model performance. The intensive experiments are conducted by
using four basic Non-IID settings and a real-world dataset. The results show
that StoCFL could obtain promising cluster results even when the number of
clusters is unknown. Based on the client clustering results, models trained
with StoCFL outperform baseline approaches in a variety of contexts.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 01:39:16 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Zeng",
"Dun",
""
],
[
"Hu",
"Xiangjing",
""
],
[
"Liu",
"Shiyu",
""
],
[
"Yu",
"Yue",
""
],
[
"Wang",
"Qifan",
""
],
[
"Xu",
"Zenglin",
""
]
] | TITLE: Stochastic Clustered Federated Learning
ABSTRACT: Federated learning is a distributed learning framework that takes full
advantage of private data samples kept on edge devices. In real-world federated
learning systems, these data samples are often decentralized and
Non-Independently Identically Distributed (Non-IID), causing divergence and
performance degradation in the federated learning process. As a new solution,
clustered federated learning groups federated clients with similar data
distributions to impair the Non-IID effects and train a better model for every
cluster. This paper proposes StoCFL, a novel clustered federated learning
approach for generic Non-IID issues. In detail, StoCFL implements a flexible
CFL framework that supports an arbitrary proportion of client participation and
newly joined clients for a varying FL system, while maintaining a great
improvement in model performance. The intensive experiments are conducted by
using four basic Non-IID settings and a real-world dataset. The results show
that StoCFL could obtain promising cluster results even when the number of
clusters is unknown. Based on the client clustering results, models trained
with StoCFL outperform baseline approaches in a variety of contexts.
|
2303.00900 | Siyuan Yan | Siyuan Yan, Jing Zhang, Nick Barnes | Transmission-Guided Bayesian Generative Model for Smoke Segmentation | Accepted by AAAI2022 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Smoke segmentation is essential to precisely localize wildfire so that it can
be extinguished in an early phase. Although deep neural networks have achieved
promising results on image segmentation tasks, they are prone to be
overconfident for smoke segmentation due to its non-rigid shape and transparent
appearance. This is caused by both knowledge level uncertainty due to limited
training data for accurate smoke segmentation and labeling level uncertainty
representing the difficulty in labeling ground-truth. To effectively model the
two types of uncertainty, we introduce a Bayesian generative model to
simultaneously estimate the posterior distribution of model parameters and its
predictions. Further, smoke images suffer from low contrast and ambiguity,
inspired by physics-based image dehazing methods, we design a
transmission-guided local coherence loss to guide the network to learn
pair-wise relationships based on pixel distance and the transmission feature.
To promote the development of this field, we also contribute a high-quality
smoke segmentation dataset, SMOKE5K, consisting of 1,400 real and 4,000
synthetic images with pixel-wise annotation. Experimental results on benchmark
testing datasets illustrate that our model achieves both accurate predictions
and reliable uncertainty maps representing model ignorance about its
prediction. Our code and dataset are publicly available at:
https://github.com/redlessme/Transmission-BVM.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 01:48:05 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yan",
"Siyuan",
""
],
[
"Zhang",
"Jing",
""
],
[
"Barnes",
"Nick",
""
]
] | TITLE: Transmission-Guided Bayesian Generative Model for Smoke Segmentation
ABSTRACT: Smoke segmentation is essential to precisely localize wildfire so that it can
be extinguished in an early phase. Although deep neural networks have achieved
promising results on image segmentation tasks, they are prone to be
overconfident for smoke segmentation due to its non-rigid shape and transparent
appearance. This is caused by both knowledge level uncertainty due to limited
training data for accurate smoke segmentation and labeling level uncertainty
representing the difficulty in labeling ground-truth. To effectively model the
two types of uncertainty, we introduce a Bayesian generative model to
simultaneously estimate the posterior distribution of model parameters and its
predictions. Further, smoke images suffer from low contrast and ambiguity,
inspired by physics-based image dehazing methods, we design a
transmission-guided local coherence loss to guide the network to learn
pair-wise relationships based on pixel distance and the transmission feature.
To promote the development of this field, we also contribute a high-quality
smoke segmentation dataset, SMOKE5K, consisting of 1,400 real and 4,000
synthetic images with pixel-wise annotation. Experimental results on benchmark
testing datasets illustrate that our model achieves both accurate predictions
and reliable uncertainty maps representing model ignorance about its
prediction. Our code and dataset are publicly available at:
https://github.com/redlessme/Transmission-BVM.
|
2303.00902 | Tugrulcan Elmas | Tu\u{g}rulcan Elmas | The Impact of Data Persistence Bias on Social Media Studies | In Proceedings of ACM WebSci23 | null | 10.1145/3578503.3583630 | null | cs.SI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Social media studies often collect data retrospectively to analyze public
opinion. Social media data may decay over time and such decay may prevent the
collection of the complete dataset. As a result, the collected dataset may
differ from the complete dataset and the study may suffer from data persistence
bias. Past research suggests that the datasets collected retrospectively are
largely representative of the original dataset in terms of textual content.
However, no study analyzed the impact of data persistence bias on social media
studies such as those focusing on controversial topics. In this study, we
analyze the data persistence and the bias it introduces on the datasets of
three types: controversial topics, trending topics, and framing of issues. We
report which topics are more likely to suffer from data persistence among these
datasets. We quantify the data persistence bias using the change in political
orientation, the presence of potentially harmful content and topics as
measures. We found that controversial datasets are more likely to suffer from
data persistence and they lean towards the political left upon recollection.
The turnout of the data that contain potentially harmful content is
significantly lower on non-controversial datasets. Overall, we found that the
topics promoted by right-aligned users are more likely to suffer from data
persistence. Account suspensions are the primary factor contributing to data
removals, if not the only one. Our results emphasize the importance of
accounting for the data persistence bias by collecting the data in real time
when the dataset employed is vulnerable to data persistence bias.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 01:51:28 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Elmas",
"Tuğrulcan",
""
]
] | TITLE: The Impact of Data Persistence Bias on Social Media Studies
ABSTRACT: Social media studies often collect data retrospectively to analyze public
opinion. Social media data may decay over time and such decay may prevent the
collection of the complete dataset. As a result, the collected dataset may
differ from the complete dataset and the study may suffer from data persistence
bias. Past research suggests that the datasets collected retrospectively are
largely representative of the original dataset in terms of textual content.
However, no study analyzed the impact of data persistence bias on social media
studies such as those focusing on controversial topics. In this study, we
analyze the data persistence and the bias it introduces on the datasets of
three types: controversial topics, trending topics, and framing of issues. We
report which topics are more likely to suffer from data persistence among these
datasets. We quantify the data persistence bias using the change in political
orientation, the presence of potentially harmful content and topics as
measures. We found that controversial datasets are more likely to suffer from
data persistence and they lean towards the political left upon recollection.
The turnout of the data that contain potentially harmful content is
significantly lower on non-controversial datasets. Overall, we found that the
topics promoted by right-aligned users are more likely to suffer from data
persistence. Account suspensions are the primary factor contributing to data
removals, if not the only one. Our results emphasize the importance of
accounting for the data persistence bias by collecting the data in real time
when the dataset employed is vulnerable to data persistence bias.
|
2303.00912 | Woojun Kim | Woojun Kim, Youngchul Sung | Parameter Sharing with Network Pruning for Scalable Multi-Agent Deep
Reinforcement Learning | null | AAMAS-2023 | null | null | cs.MA cs.AI | http://creativecommons.org/licenses/by/4.0/ | Handling the problem of scalability is one of the essential issues for
multi-agent reinforcement learning (MARL) algorithms to be applied to
real-world problems typically involving massively many agents. For this,
parameter sharing across multiple agents has widely been used since it reduces
the training time by decreasing the number of parameters and increasing the
sample efficiency. However, using the same parameters across agents limits the
representational capacity of the joint policy and consequently, the performance
can be degraded in multi-agent tasks that require different behaviors for
different agents. In this paper, we propose a simple method that adopts
structured pruning for a deep neural network to increase the representational
capacity of the joint policy without introducing additional parameters. We
evaluate the proposed method on several benchmark tasks, and numerical results
show that the proposed method significantly outperforms other parameter-sharing
methods.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 02:17:14 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kim",
"Woojun",
""
],
[
"Sung",
"Youngchul",
""
]
] | TITLE: Parameter Sharing with Network Pruning for Scalable Multi-Agent Deep
Reinforcement Learning
ABSTRACT: Handling the problem of scalability is one of the essential issues for
multi-agent reinforcement learning (MARL) algorithms to be applied to
real-world problems typically involving massively many agents. For this,
parameter sharing across multiple agents has widely been used since it reduces
the training time by decreasing the number of parameters and increasing the
sample efficiency. However, using the same parameters across agents limits the
representational capacity of the joint policy and consequently, the performance
can be degraded in multi-agent tasks that require different behaviors for
different agents. In this paper, we propose a simple method that adopts
structured pruning for a deep neural network to increase the representational
capacity of the joint policy without introducing additional parameters. We
evaluate the proposed method on several benchmark tasks, and numerical results
show that the proposed method significantly outperforms other parameter-sharing
methods.
|
2303.00918 | Jaehyun Nam | Jaehyun Nam, Jihoon Tack, Kyungmin Lee, Hankook Lee, Jinwoo Shin | STUNT: Few-shot Tabular Learning with Self-generated Tasks from
Unlabeled Tables | ICLR 2023 (Spotlight) | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning with few labeled tabular samples is often an essential requirement
for industrial machine learning applications as varieties of tabular data
suffer from high annotation costs or have difficulties in collecting new
samples for novel tasks. Despite the utter importance, such a problem is quite
under-explored in the field of tabular learning, and existing few-shot learning
schemes from other domains are not straightforward to apply, mainly due to the
heterogeneous characteristics of tabular data. In this paper, we propose a
simple yet effective framework for few-shot semi-supervised tabular learning,
coined Self-generated Tasks from UNlabeled Tables (STUNT). Our key idea is to
self-generate diverse few-shot tasks by treating randomly chosen columns as a
target label. We then employ a meta-learning scheme to learn generalizable
knowledge with the constructed tasks. Moreover, we introduce an unsupervised
validation scheme for hyperparameter search (and early stopping) by generating
a pseudo-validation set using STUNT from unlabeled data. Our experimental
results demonstrate that our simple framework brings significant performance
gain under various tabular few-shot learning benchmarks, compared to prior
semi- and self-supervised baselines. Code is available at
https://github.com/jaehyun513/STUNT.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 02:37:54 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Nam",
"Jaehyun",
""
],
[
"Tack",
"Jihoon",
""
],
[
"Lee",
"Kyungmin",
""
],
[
"Lee",
"Hankook",
""
],
[
"Shin",
"Jinwoo",
""
]
] | TITLE: STUNT: Few-shot Tabular Learning with Self-generated Tasks from
Unlabeled Tables
ABSTRACT: Learning with few labeled tabular samples is often an essential requirement
for industrial machine learning applications as varieties of tabular data
suffer from high annotation costs or have difficulties in collecting new
samples for novel tasks. Despite the utter importance, such a problem is quite
under-explored in the field of tabular learning, and existing few-shot learning
schemes from other domains are not straightforward to apply, mainly due to the
heterogeneous characteristics of tabular data. In this paper, we propose a
simple yet effective framework for few-shot semi-supervised tabular learning,
coined Self-generated Tasks from UNlabeled Tables (STUNT). Our key idea is to
self-generate diverse few-shot tasks by treating randomly chosen columns as a
target label. We then employ a meta-learning scheme to learn generalizable
knowledge with the constructed tasks. Moreover, we introduce an unsupervised
validation scheme for hyperparameter search (and early stopping) by generating
a pseudo-validation set using STUNT from unlabeled data. Our experimental
results demonstrate that our simple framework brings significant performance
gain under various tabular few-shot learning benchmarks, compared to prior
semi- and self-supervised baselines. Code is available at
https://github.com/jaehyun513/STUNT.
|
2303.00922 | Tarik A. Rashid | Mohammad Khishe, Mokhtar Mohammadi, Tarik A. Rashid, Hoger Mahmud,
Seyedali Mirjalili | Evolving Deep Neural Network by Customized Moth Flame Optimization
Algorithm for Underwater Targets Recognition | 16 pages | null | null | null | cs.NE | http://creativecommons.org/licenses/by/4.0/ | This chapter proposes using the Moth Flame Optimization (MFO) algorithm for
finetuning a Deep Neural Network to recognize different underwater sonar
datasets. Same as other models evolved by metaheuristic algorithms, premature
convergence, trapping in local minima, and failure to converge in a reasonable
time are three defects MFO confronts in solving problems with high-dimension
search space. Spiral flying is the key component of the MFO as it determines
how the moths adjust their positions in relation to flames; thereby, the shape
of spiral motions can regulate the transition behavior between the exploration
and exploitation phases. Therefore, this chapter investigates the efficiency of
seven spiral motions with different curvatures and slopes in the performance of
the MFO, especially for underwater target classification tasks. To assess the
performance of the customized model, in addition to benchmark Sejnowski &
Gorman's dataset, two experimental sonar datasets, i.e., the passive sonar and
active datasets, are exploited. The results of MFO and its modifications are
compared to four novel nature-inspired algorithms, including Heap-Based
Optimizer (HBO), Chimp Optimization Algorithm (ChOA), Ant Lion Optimization
(ALO), Stochastic Fractals Search (SFS), as well as the classic Particle Swarm
Optimization (PSO). The results confirm that the customized MFO shows better
performance than the other state-of-the-art models so that the classification
rates are increased 1.5979, 0.9985, and 2.0879 for Sejnowski & Gorman, passive,
and active datasets, respectively. The results also approve that time
complexity is not significantly increased by using different spiral motions.
| [
{
"version": "v1",
"created": "Thu, 16 Feb 2023 18:18:18 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Khishe",
"Mohammad",
""
],
[
"Mohammadi",
"Mokhtar",
""
],
[
"Rashid",
"Tarik A.",
""
],
[
"Mahmud",
"Hoger",
""
],
[
"Mirjalili",
"Seyedali",
""
]
] | TITLE: Evolving Deep Neural Network by Customized Moth Flame Optimization
Algorithm for Underwater Targets Recognition
ABSTRACT: This chapter proposes using the Moth Flame Optimization (MFO) algorithm for
finetuning a Deep Neural Network to recognize different underwater sonar
datasets. Same as other models evolved by metaheuristic algorithms, premature
convergence, trapping in local minima, and failure to converge in a reasonable
time are three defects MFO confronts in solving problems with high-dimension
search space. Spiral flying is the key component of the MFO as it determines
how the moths adjust their positions in relation to flames; thereby, the shape
of spiral motions can regulate the transition behavior between the exploration
and exploitation phases. Therefore, this chapter investigates the efficiency of
seven spiral motions with different curvatures and slopes in the performance of
the MFO, especially for underwater target classification tasks. To assess the
performance of the customized model, in addition to benchmark Sejnowski &
Gorman's dataset, two experimental sonar datasets, i.e., the passive sonar and
active datasets, are exploited. The results of MFO and its modifications are
compared to four novel nature-inspired algorithms, including Heap-Based
Optimizer (HBO), Chimp Optimization Algorithm (ChOA), Ant Lion Optimization
(ALO), Stochastic Fractals Search (SFS), as well as the classic Particle Swarm
Optimization (PSO). The results confirm that the customized MFO shows better
performance than the other state-of-the-art models so that the classification
rates are increased 1.5979, 0.9985, and 2.0879 for Sejnowski & Gorman, passive,
and active datasets, respectively. The results also approve that time
complexity is not significantly increased by using different spiral motions.
|
2303.00923 | Mir Tafseer Nayeem | Mir Tafseer Nayeem, Davood Rafiei | On the Role of Reviewer Expertise in Temporal Review Helpfulness
Prediction | Accepted to Findings of EACL 2023 | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | Helpful reviews have been essential for the success of e-commerce services,
as they help customers make quick purchase decisions and benefit the merchants
in their sales. While many reviews are informative, others provide little value
and may contain spam, excessive appraisal, or unexpected biases. With the large
volume of reviews and their uneven quality, the problem of detecting helpful
reviews has drawn much attention lately. Existing methods for identifying
helpful reviews primarily focus on review text and ignore the two key factors
of (1) who post the reviews and (2) when the reviews are posted. Moreover, the
helpfulness votes suffer from scarcity for less popular products and recently
submitted (a.k.a., cold-start) reviews. To address these challenges, we
introduce a dataset and develop a model that integrates the reviewer's
expertise, derived from the past review history of the reviewers, and the
temporal dynamics of the reviews to automatically assess review helpfulness. We
conduct experiments on our dataset to demonstrate the effectiveness of
incorporating these factors and report improved results compared to several
well-established baselines.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 23:41:22 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Nayeem",
"Mir Tafseer",
""
],
[
"Rafiei",
"Davood",
""
]
] | TITLE: On the Role of Reviewer Expertise in Temporal Review Helpfulness
Prediction
ABSTRACT: Helpful reviews have been essential for the success of e-commerce services,
as they help customers make quick purchase decisions and benefit the merchants
in their sales. While many reviews are informative, others provide little value
and may contain spam, excessive appraisal, or unexpected biases. With the large
volume of reviews and their uneven quality, the problem of detecting helpful
reviews has drawn much attention lately. Existing methods for identifying
helpful reviews primarily focus on review text and ignore the two key factors
of (1) who post the reviews and (2) when the reviews are posted. Moreover, the
helpfulness votes suffer from scarcity for less popular products and recently
submitted (a.k.a., cold-start) reviews. To address these challenges, we
introduce a dataset and develop a model that integrates the reviewer's
expertise, derived from the past review history of the reviewers, and the
temporal dynamics of the reviews to automatically assess review helpfulness. We
conduct experiments on our dataset to demonstrate the effectiveness of
incorporating these factors and report improved results compared to several
well-established baselines.
|
2303.00928 | Soumi Chaki | Soumi Chaki, David Weinberg, and Ayca \"Ozcelikkale | Communication Trade-offs in Federated Learning of Spiking Neural
Networks | 5 pages, 7 figures, 1 algorithm, 1 table | null | null | null | cs.LG cs.NE eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spiking Neural Networks (SNNs) are biologically inspired alternatives to
conventional Artificial Neural Networks (ANNs). Despite promising preliminary
results, the trade-offs in the training of SNNs in a distributed scheme are not
well understood. Here, we consider SNNs in a federated learning setting where a
high-quality global model is created by aggregating multiple local models from
the clients without sharing any data. We investigate federated learning for
training multiple SNNs at clients when two mechanisms reduce the uplink
communication cost: i) random masking of the model updates sent from the
clients to the server; and ii) client dropouts where some clients do not send
their updates to the server. We evaluated the performance of the SNNs using a
subset of the Spiking Heidelberg digits (SHD) dataset. The results show that a
trade-off between the random masking and the client drop probabilities is
crucial to obtain a satisfactory performance for a fixed number of clients.
| [
{
"version": "v1",
"created": "Mon, 27 Feb 2023 19:12:03 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chaki",
"Soumi",
""
],
[
"Weinberg",
"David",
""
],
[
"Özcelikkale",
"Ayca",
""
]
] | TITLE: Communication Trade-offs in Federated Learning of Spiking Neural
Networks
ABSTRACT: Spiking Neural Networks (SNNs) are biologically inspired alternatives to
conventional Artificial Neural Networks (ANNs). Despite promising preliminary
results, the trade-offs in the training of SNNs in a distributed scheme are not
well understood. Here, we consider SNNs in a federated learning setting where a
high-quality global model is created by aggregating multiple local models from
the clients without sharing any data. We investigate federated learning for
training multiple SNNs at clients when two mechanisms reduce the uplink
communication cost: i) random masking of the model updates sent from the
clients to the server; and ii) client dropouts where some clients do not send
their updates to the server. We evaluated the performance of the SNNs using a
subset of the Spiking Heidelberg digits (SHD) dataset. The results show that a
trade-off between the random masking and the client drop probabilities is
crucial to obtain a satisfactory performance for a fixed number of clients.
|
2303.00942 | Bo Zhou | Bo Zhou, Yingda Xia, Jiawen Yao, Le Lu, Jingren Zhou, Chi Liu, James
S. Duncan, Ling Zhang | Meta-information-aware Dual-path Transformer for Differential Diagnosis
of Multi-type Pancreatic Lesions in Multi-phase CT | Accepted at Information Processing in Medical Imaging (IPMI 2023) | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pancreatic cancer is one of the leading causes of cancer-related death.
Accurate detection, segmentation, and differential diagnosis of the full
taxonomy of pancreatic lesions, i.e., normal, seven major types of lesions, and
other lesions, is critical to aid the clinical decision-making of patient
management and treatment. However, existing works focus on segmentation and
classification for very specific lesion types (PDAC) or groups. Moreover, none
of the previous work considers using lesion prevalence-related non-imaging
patient information to assist the differential diagnosis. To this end, we
develop a meta-information-aware dual-path transformer and exploit the
feasibility of classification and segmentation of the full taxonomy of
pancreatic lesions. Specifically, the proposed method consists of a CNN-based
segmentation path (S-path) and a transformer-based classification path
(C-path). The S-path focuses on initial feature extraction by semantic
segmentation using a UNet-based network. The C-path utilizes both the extracted
features and meta-information for patient-level classification based on stacks
of dual-path transformer blocks that enhance the modeling of global contextual
information. A large-scale multi-phase CT dataset of 3,096 patients with
pathology-confirmed pancreatic lesion class labels, voxel-wise manual
annotations of lesions from radiologists, and patient meta-information, was
collected for training and evaluations. Our results show that our method can
enable accurate classification and segmentation of the full taxonomy of
pancreatic lesions, approaching the accuracy of the radiologist's report and
significantly outperforming previous baselines. Results also show that adding
the common meta-information, i.e., gender and age, can boost the model's
performance, thus demonstrating the importance of meta-information for aiding
pancreatic disease diagnosis.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 03:34:28 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Zhou",
"Bo",
""
],
[
"Xia",
"Yingda",
""
],
[
"Yao",
"Jiawen",
""
],
[
"Lu",
"Le",
""
],
[
"Zhou",
"Jingren",
""
],
[
"Liu",
"Chi",
""
],
[
"Duncan",
"James S.",
""
],
[
"Zhang",
"Ling",
""
]
] | TITLE: Meta-information-aware Dual-path Transformer for Differential Diagnosis
of Multi-type Pancreatic Lesions in Multi-phase CT
ABSTRACT: Pancreatic cancer is one of the leading causes of cancer-related death.
Accurate detection, segmentation, and differential diagnosis of the full
taxonomy of pancreatic lesions, i.e., normal, seven major types of lesions, and
other lesions, is critical to aid the clinical decision-making of patient
management and treatment. However, existing works focus on segmentation and
classification for very specific lesion types (PDAC) or groups. Moreover, none
of the previous work considers using lesion prevalence-related non-imaging
patient information to assist the differential diagnosis. To this end, we
develop a meta-information-aware dual-path transformer and exploit the
feasibility of classification and segmentation of the full taxonomy of
pancreatic lesions. Specifically, the proposed method consists of a CNN-based
segmentation path (S-path) and a transformer-based classification path
(C-path). The S-path focuses on initial feature extraction by semantic
segmentation using a UNet-based network. The C-path utilizes both the extracted
features and meta-information for patient-level classification based on stacks
of dual-path transformer blocks that enhance the modeling of global contextual
information. A large-scale multi-phase CT dataset of 3,096 patients with
pathology-confirmed pancreatic lesion class labels, voxel-wise manual
annotations of lesions from radiologists, and patient meta-information, was
collected for training and evaluations. Our results show that our method can
enable accurate classification and segmentation of the full taxonomy of
pancreatic lesions, approaching the accuracy of the radiologist's report and
significantly outperforming previous baselines. Results also show that adding
the common meta-information, i.e., gender and age, can boost the model's
performance, thus demonstrating the importance of meta-information for aiding
pancreatic disease diagnosis.
|
2303.00977 | Chihiro Noguchi | Chihiro Noguchi, Toshihiro Tanizawa | Ego-Vehicle Action Recognition based on Semi-Supervised Contrastive
Learning | 19 pages, 17 figures | Proceedings of the IEEE/CVF Winter Conference on Applications of
Computer Vision. 2023. p. 5988-5998 | 10.1109/WACV56688.2023.00593 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, many automobiles have been equipped with cameras, which have
accumulated an enormous amount of video footage of driving scenes. Autonomous
driving demands the highest level of safety, for which even unimaginably rare
driving scenes have to be collected in training data to improve the recognition
accuracy for specific scenes. However, it is prohibitively costly to find very
few specific scenes from an enormous amount of videos. In this article, we show
that proper video-to-video distances can be defined by focusing on ego-vehicle
actions. It is well known that existing methods based on supervised learning
cannot handle videos that do not fall into predefined classes, though they work
well in defining video-to-video distances in the embedding space between
labeled videos. To tackle this problem, we propose a method based on
semi-supervised contrastive learning. We consider two related but distinct
contrastive learning: standard graph contrastive learning and our proposed
SOIA-based contrastive learning. We observe that the latter approach can
provide more sensible video-to-video distances between unlabeled videos. Next,
the effectiveness of our method is quantified by evaluating the classification
performance of the ego-vehicle action recognition using HDD dataset, which
shows that our method including unlabeled data in training significantly
outperforms the existing methods using only labeled data in training.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 05:19:31 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Noguchi",
"Chihiro",
""
],
[
"Tanizawa",
"Toshihiro",
""
]
] | TITLE: Ego-Vehicle Action Recognition based on Semi-Supervised Contrastive
Learning
ABSTRACT: In recent years, many automobiles have been equipped with cameras, which have
accumulated an enormous amount of video footage of driving scenes. Autonomous
driving demands the highest level of safety, for which even unimaginably rare
driving scenes have to be collected in training data to improve the recognition
accuracy for specific scenes. However, it is prohibitively costly to find very
few specific scenes from an enormous amount of videos. In this article, we show
that proper video-to-video distances can be defined by focusing on ego-vehicle
actions. It is well known that existing methods based on supervised learning
cannot handle videos that do not fall into predefined classes, though they work
well in defining video-to-video distances in the embedding space between
labeled videos. To tackle this problem, we propose a method based on
semi-supervised contrastive learning. We consider two related but distinct
contrastive learning: standard graph contrastive learning and our proposed
SOIA-based contrastive learning. We observe that the latter approach can
provide more sensible video-to-video distances between unlabeled videos. Next,
the effectiveness of our method is quantified by evaluating the classification
performance of the ego-vehicle action recognition using HDD dataset, which
shows that our method including unlabeled data in training significantly
outperforms the existing methods using only labeled data in training.
|
2303.00978 | Kohei Matsuura | Kohei Matsuura, Takanori Ashihara, Takafumi Moriya, Tomohiro Tanaka,
Atsunori Ogawa, Marc Delcroix, Ryo Masumura | Leveraging Large Text Corpora for End-to-End Speech Summarization | Accepted to ICASSP 2023 | null | null | null | cs.CL eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | End-to-end speech summarization (E2E SSum) is a technique to directly
generate summary sentences from speech. Compared with the cascade approach,
which combines automatic speech recognition (ASR) and text summarization
models, the E2E approach is more promising because it mitigates ASR errors,
incorporates nonverbal information, and simplifies the overall system. However,
since collecting a large amount of paired data (i.e., speech and summary) is
difficult, the training data is usually insufficient to train a robust E2E SSum
system. In this paper, we present two novel methods that leverage a large
amount of external text summarization data for E2E SSum training. The first
technique is to utilize a text-to-speech (TTS) system to generate synthesized
speech, which is used for E2E SSum training with the text summary. The second
is a TTS-free method that directly inputs phoneme sequence instead of
synthesized speech to the E2E SSum model. Experiments show that our proposed
TTS- and phoneme-based methods improve several metrics on the How2 dataset. In
particular, our best system outperforms a previous state-of-the-art one by a
large margin (i.e., METEOR score improvements of more than 6 points). To the
best of our knowledge, this is the first work to use external language
resources for E2E SSum. Moreover, we report a detailed analysis of the How2
dataset to confirm the validity of our proposed E2E SSum system.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 05:19:49 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Matsuura",
"Kohei",
""
],
[
"Ashihara",
"Takanori",
""
],
[
"Moriya",
"Takafumi",
""
],
[
"Tanaka",
"Tomohiro",
""
],
[
"Ogawa",
"Atsunori",
""
],
[
"Delcroix",
"Marc",
""
],
[
"Masumura",
"Ryo",
""
]
] | TITLE: Leveraging Large Text Corpora for End-to-End Speech Summarization
ABSTRACT: End-to-end speech summarization (E2E SSum) is a technique to directly
generate summary sentences from speech. Compared with the cascade approach,
which combines automatic speech recognition (ASR) and text summarization
models, the E2E approach is more promising because it mitigates ASR errors,
incorporates nonverbal information, and simplifies the overall system. However,
since collecting a large amount of paired data (i.e., speech and summary) is
difficult, the training data is usually insufficient to train a robust E2E SSum
system. In this paper, we present two novel methods that leverage a large
amount of external text summarization data for E2E SSum training. The first
technique is to utilize a text-to-speech (TTS) system to generate synthesized
speech, which is used for E2E SSum training with the text summary. The second
is a TTS-free method that directly inputs phoneme sequence instead of
synthesized speech to the E2E SSum model. Experiments show that our proposed
TTS- and phoneme-based methods improve several metrics on the How2 dataset. In
particular, our best system outperforms a previous state-of-the-art one by a
large margin (i.e., METEOR score improvements of more than 6 points). To the
best of our knowledge, this is the first work to use external language
resources for E2E SSum. Moreover, we report a detailed analysis of the How2
dataset to confirm the validity of our proposed E2E SSum system.
|
2303.00995 | Wei Wei | Mengru Chen and Chao Huang and Lianghao Xia and Wei Wei and Yong Xu
and Ronghua Luo | Heterogeneous Graph Contrastive Learning for Recommendation | This paper has been published as a full paper at WSDM 2023 | null | 10.1145/3539597.3570484 | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph Neural Networks (GNNs) have become powerful tools in modeling
graph-structured data in recommender systems. However, real-life recommendation
scenarios usually involve heterogeneous relationships (e.g., social-aware user
influence, knowledge-aware item dependency) which contains fruitful information
to enhance the user preference learning. In this paper, we study the problem of
heterogeneous graph-enhanced relational learning for recommendation. Recently,
contrastive self-supervised learning has become successful in recommendation.
In light of this, we propose a Heterogeneous Graph Contrastive Learning (HGCL),
which is able to incorporate heterogeneous relational semantics into the
user-item interaction modeling with contrastive learning-enhanced knowledge
transfer across different views. However, the influence of heterogeneous side
information on interactions may vary by users and items. To move this idea
forward, we enhance our heterogeneous graph contrastive learning with meta
networks to allow the personalized knowledge transformer with adaptive
contrastive augmentation. The experimental results on three real-world datasets
demonstrate the superiority of HGCL over state-of-the-art recommendation
methods. Through ablation study, key components in HGCL method are validated to
benefit the recommendation performance improvement. The source code of the
model implementation is available at the link https://github.com/HKUDS/HGCL.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 06:07:22 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chen",
"Mengru",
""
],
[
"Huang",
"Chao",
""
],
[
"Xia",
"Lianghao",
""
],
[
"Wei",
"Wei",
""
],
[
"Xu",
"Yong",
""
],
[
"Luo",
"Ronghua",
""
]
] | TITLE: Heterogeneous Graph Contrastive Learning for Recommendation
ABSTRACT: Graph Neural Networks (GNNs) have become powerful tools in modeling
graph-structured data in recommender systems. However, real-life recommendation
scenarios usually involve heterogeneous relationships (e.g., social-aware user
influence, knowledge-aware item dependency) which contains fruitful information
to enhance the user preference learning. In this paper, we study the problem of
heterogeneous graph-enhanced relational learning for recommendation. Recently,
contrastive self-supervised learning has become successful in recommendation.
In light of this, we propose a Heterogeneous Graph Contrastive Learning (HGCL),
which is able to incorporate heterogeneous relational semantics into the
user-item interaction modeling with contrastive learning-enhanced knowledge
transfer across different views. However, the influence of heterogeneous side
information on interactions may vary by users and items. To move this idea
forward, we enhance our heterogeneous graph contrastive learning with meta
networks to allow the personalized knowledge transformer with adaptive
contrastive augmentation. The experimental results on three real-world datasets
demonstrate the superiority of HGCL over state-of-the-art recommendation
methods. Through ablation study, key components in HGCL method are validated to
benefit the recommendation performance improvement. The source code of the
model implementation is available at the link https://github.com/HKUDS/HGCL.
|
2303.00996 | Huiwon Jang | Huiwon Jang, Hankook Lee, Jinwoo Shin | Unsupervised Meta-Learning via Few-shot Pseudo-supervised Contrastive
Learning | Accepted to ICLR 2023 (Spotlight). The first two authors contributed
equally. The code is available at https://github.com/alinlab/PsCo | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Unsupervised meta-learning aims to learn generalizable knowledge across a
distribution of tasks constructed from unlabeled data. Here, the main challenge
is how to construct diverse tasks for meta-learning without label information;
recent works have proposed to create, e.g., pseudo-labeling via pretrained
representations or creating synthetic samples via generative models. However,
such a task construction strategy is fundamentally limited due to heavy
reliance on the immutable pseudo-labels during meta-learning and the quality of
the representations or the generated samples. To overcome the limitations, we
propose a simple yet effective unsupervised meta-learning framework, coined
Pseudo-supervised Contrast (PsCo), for few-shot classification. We are inspired
by the recent self-supervised learning literature; PsCo utilizes a momentum
network and a queue of previous batches to improve pseudo-labeling and
construct diverse tasks in a progressive manner. Our extensive experiments
demonstrate that PsCo outperforms existing unsupervised meta-learning methods
under various in-domain and cross-domain few-shot classification benchmarks. We
also validate that PsCo is easily scalable to a large-scale benchmark, while
recent prior-art meta-schemes are not.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 06:10:13 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Jang",
"Huiwon",
""
],
[
"Lee",
"Hankook",
""
],
[
"Shin",
"Jinwoo",
""
]
] | TITLE: Unsupervised Meta-Learning via Few-shot Pseudo-supervised Contrastive
Learning
ABSTRACT: Unsupervised meta-learning aims to learn generalizable knowledge across a
distribution of tasks constructed from unlabeled data. Here, the main challenge
is how to construct diverse tasks for meta-learning without label information;
recent works have proposed to create, e.g., pseudo-labeling via pretrained
representations or creating synthetic samples via generative models. However,
such a task construction strategy is fundamentally limited due to heavy
reliance on the immutable pseudo-labels during meta-learning and the quality of
the representations or the generated samples. To overcome the limitations, we
propose a simple yet effective unsupervised meta-learning framework, coined
Pseudo-supervised Contrast (PsCo), for few-shot classification. We are inspired
by the recent self-supervised learning literature; PsCo utilizes a momentum
network and a queue of previous batches to improve pseudo-labeling and
construct diverse tasks in a progressive manner. Our extensive experiments
demonstrate that PsCo outperforms existing unsupervised meta-learning methods
under various in-domain and cross-domain few-shot classification benchmarks. We
also validate that PsCo is easily scalable to a large-scale benchmark, while
recent prior-art meta-schemes are not.
|
2303.01000 | Yuval Kirstain | Yuval Kirstain, Omer Levy, Adam Polyak | X&Fuse: Fusing Visual Information in Text-to-Image Generation | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | We introduce X&Fuse, a general approach for conditioning on visual
information when generating images from text. We demonstrate the potential of
X&Fuse in three different text-to-image generation scenarios. (i) When a bank
of images is available, we retrieve and condition on a related image
(Retrieve&Fuse), resulting in significant improvements on the MS-COCO
benchmark, gaining a state-of-the-art FID score of 6.65 in zero-shot settings.
(ii) When cropped-object images are at hand, we utilize them and perform
subject-driven generation (Crop&Fuse), outperforming the textual inversion
method while being more than x100 faster. (iii) Having oracle access to the
image scene (Scene&Fuse), allows us to achieve an FID score of 5.03 on MS-COCO
in zero-shot settings. Our experiments indicate that X&Fuse is an effective,
easy-to-adapt, simple, and general approach for scenarios in which the model
may benefit from additional visual information.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 06:33:33 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kirstain",
"Yuval",
""
],
[
"Levy",
"Omer",
""
],
[
"Polyak",
"Adam",
""
]
] | TITLE: X&Fuse: Fusing Visual Information in Text-to-Image Generation
ABSTRACT: We introduce X&Fuse, a general approach for conditioning on visual
information when generating images from text. We demonstrate the potential of
X&Fuse in three different text-to-image generation scenarios. (i) When a bank
of images is available, we retrieve and condition on a related image
(Retrieve&Fuse), resulting in significant improvements on the MS-COCO
benchmark, gaining a state-of-the-art FID score of 6.65 in zero-shot settings.
(ii) When cropped-object images are at hand, we utilize them and perform
subject-driven generation (Crop&Fuse), outperforming the textual inversion
method while being more than x100 faster. (iii) Having oracle access to the
image scene (Scene&Fuse), allows us to achieve an FID score of 5.03 on MS-COCO
in zero-shot settings. Our experiments indicate that X&Fuse is an effective,
easy-to-adapt, simple, and general approach for scenarios in which the model
may benefit from additional visual information.
|
2303.01021 | Yair Meidan | Yair Meidan, Dan Avraham, Hanan Libhaber, Asaf Shabtai | CADeSH: Collaborative Anomaly Detection for Smart Homes | in IEEE Internet of Things Journal, 2022 | null | 10.1109/JIOT.2022.3194813 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Although home IoT (Internet of Things) devices are typically plain and task
oriented, the context of their daily use may affect their traffic patterns. For
this reason, anomaly-based intrusion detection systems tend to suffer from a
high false positive rate (FPR). To overcome this, we propose a two-step
collaborative anomaly detection method which first uses an autoencoder to
differentiate frequent (`benign') and infrequent (possibly `malicious') traffic
flows. Clustering is then used to analyze only the infrequent flows and
classify them as either known ('rare yet benign') or unknown (`malicious'). Our
method is collaborative, in that (1) normal behaviors are characterized more
robustly, as they take into account a variety of user interactions and network
topologies, and (2) several features are computed based on a pool of identical
devices rather than just the inspected device.
We evaluated our method empirically, using 21 days of real-world traffic data
that emanated from eight identical IoT devices deployed on various networks,
one of which was located in our controlled lab where we implemented two popular
IoT-related cyber-attacks. Our collaborative anomaly detection method achieved
a macro-average area under the precision-recall curve of 0.841, an F1 score of
0.929, and an FPR of only 0.014. These promising results were obtained by using
labeled traffic data from our lab as the test set, while training the models on
the traffic of devices deployed outside the lab, and thus demonstrate a high
level of generalizability. In addition to its high generalizability and
promising performance, our proposed method also offers benefits such as privacy
preservation, resource savings, and model poisoning mitigation. On top of that,
as a contribution to the scientific community, our novel dataset is available
online.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 07:22:26 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Meidan",
"Yair",
""
],
[
"Avraham",
"Dan",
""
],
[
"Libhaber",
"Hanan",
""
],
[
"Shabtai",
"Asaf",
""
]
] | TITLE: CADeSH: Collaborative Anomaly Detection for Smart Homes
ABSTRACT: Although home IoT (Internet of Things) devices are typically plain and task
oriented, the context of their daily use may affect their traffic patterns. For
this reason, anomaly-based intrusion detection systems tend to suffer from a
high false positive rate (FPR). To overcome this, we propose a two-step
collaborative anomaly detection method which first uses an autoencoder to
differentiate frequent (`benign') and infrequent (possibly `malicious') traffic
flows. Clustering is then used to analyze only the infrequent flows and
classify them as either known ('rare yet benign') or unknown (`malicious'). Our
method is collaborative, in that (1) normal behaviors are characterized more
robustly, as they take into account a variety of user interactions and network
topologies, and (2) several features are computed based on a pool of identical
devices rather than just the inspected device.
We evaluated our method empirically, using 21 days of real-world traffic data
that emanated from eight identical IoT devices deployed on various networks,
one of which was located in our controlled lab where we implemented two popular
IoT-related cyber-attacks. Our collaborative anomaly detection method achieved
a macro-average area under the precision-recall curve of 0.841, an F1 score of
0.929, and an FPR of only 0.014. These promising results were obtained by using
labeled traffic data from our lab as the test set, while training the models on
the traffic of devices deployed outside the lab, and thus demonstrate a high
level of generalizability. In addition to its high generalizability and
promising performance, our proposed method also offers benefits such as privacy
preservation, resource savings, and model poisoning mitigation. On top of that,
as a contribution to the scientific community, our novel dataset is available
online.
|
2303.01028 | Deyu Bo | Deyu Bo and Chuan Shi and Lele Wang and Renjie Liao | Specformer: Spectral Graph Neural Networks Meet Transformers | ICLR 2023 | null | null | null | cs.LG cs.AI cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spectral graph neural networks (GNNs) learn graph representations via
spectral-domain graph convolutions. However, most existing spectral graph
filters are scalar-to-scalar functions, i.e., mapping a single eigenvalue to a
single filtered value, thus ignoring the global pattern of the spectrum.
Furthermore, these filters are often constructed based on some fixed-order
polynomials, which have limited expressiveness and flexibility. To tackle these
issues, we introduce Specformer, which effectively encodes the set of all
eigenvalues and performs self-attention in the spectral domain, leading to a
learnable set-to-set spectral filter. We also design a decoder with learnable
bases to enable non-local graph convolution. Importantly, Specformer is
equivariant to permutation. By stacking multiple Specformer layers, one can
build a powerful spectral GNN. On synthetic datasets, we show that our
Specformer can better recover ground-truth spectral filters than other spectral
GNNs. Extensive experiments of both node-level and graph-level tasks on
real-world graph datasets show that our Specformer outperforms state-of-the-art
GNNs and learns meaningful spectrum patterns. Code and data are available at
https://github.com/bdy9527/Specformer.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 07:36:23 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Bo",
"Deyu",
""
],
[
"Shi",
"Chuan",
""
],
[
"Wang",
"Lele",
""
],
[
"Liao",
"Renjie",
""
]
] | TITLE: Specformer: Spectral Graph Neural Networks Meet Transformers
ABSTRACT: Spectral graph neural networks (GNNs) learn graph representations via
spectral-domain graph convolutions. However, most existing spectral graph
filters are scalar-to-scalar functions, i.e., mapping a single eigenvalue to a
single filtered value, thus ignoring the global pattern of the spectrum.
Furthermore, these filters are often constructed based on some fixed-order
polynomials, which have limited expressiveness and flexibility. To tackle these
issues, we introduce Specformer, which effectively encodes the set of all
eigenvalues and performs self-attention in the spectral domain, leading to a
learnable set-to-set spectral filter. We also design a decoder with learnable
bases to enable non-local graph convolution. Importantly, Specformer is
equivariant to permutation. By stacking multiple Specformer layers, one can
build a powerful spectral GNN. On synthetic datasets, we show that our
Specformer can better recover ground-truth spectral filters than other spectral
GNNs. Extensive experiments of both node-level and graph-level tasks on
real-world graph datasets show that our Specformer outperforms state-of-the-art
GNNs and learns meaningful spectrum patterns. Code and data are available at
https://github.com/bdy9527/Specformer.
|
2303.01030 | Kai Zhao | Qiyu Kang, Kai Zhao, Yang Song, Sijie Wang, Rui She, and Wee Peng Tay | Node Embedding from Hamiltonian Information Propagation in Graph Neural
Networks | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph neural networks (GNNs) have achieved success in various inference tasks
on graph-structured data. However, common challenges faced by many GNNs in the
literature include the problem of graph node embedding under various geometries
and the over-smoothing problem. To address these issues, we propose a novel
graph information propagation strategy called Hamiltonian Dynamic GNN (HDG)
that uses a Hamiltonian mechanics approach to learn node embeddings in a graph.
The Hamiltonian energy function in HDG is learnable and can adapt to the
underlying geometry of any given graph dataset. We demonstrate the ability of
HDG to automatically learn the underlying geometry of graph datasets, even
those with complex and mixed geometries, through comprehensive evaluations
against state-of-the-art baselines on various downstream tasks. We also verify
that HDG is stable against small perturbations and can mitigate the
over-smoothing problem when stacking many layers.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 07:40:40 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Kang",
"Qiyu",
""
],
[
"Zhao",
"Kai",
""
],
[
"Song",
"Yang",
""
],
[
"Wang",
"Sijie",
""
],
[
"She",
"Rui",
""
],
[
"Tay",
"Wee Peng",
""
]
] | TITLE: Node Embedding from Hamiltonian Information Propagation in Graph Neural
Networks
ABSTRACT: Graph neural networks (GNNs) have achieved success in various inference tasks
on graph-structured data. However, common challenges faced by many GNNs in the
literature include the problem of graph node embedding under various geometries
and the over-smoothing problem. To address these issues, we propose a novel
graph information propagation strategy called Hamiltonian Dynamic GNN (HDG)
that uses a Hamiltonian mechanics approach to learn node embeddings in a graph.
The Hamiltonian energy function in HDG is learnable and can adapt to the
underlying geometry of any given graph dataset. We demonstrate the ability of
HDG to automatically learn the underlying geometry of graph datasets, even
those with complex and mixed geometries, through comprehensive evaluations
against state-of-the-art baselines on various downstream tasks. We also verify
that HDG is stable against small perturbations and can mitigate the
over-smoothing problem when stacking many layers.
|
2303.01034 | Heejeong Choi | Heejeong Choi, Pilsung Kang | Multi-Task Self-Supervised Time-Series Representation Learning | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Time-series representation learning can extract representations from data
with temporal dynamics and sparse labels. When labeled data are sparse but
unlabeled data are abundant, contrastive learning, i.e., a framework to learn a
latent space where similar samples are close to each other while dissimilar
ones are far from each other, has shown outstanding performance. This strategy
can encourage varied consistency of time-series representations depending on
the positive pair selection and contrastive loss. We propose a new time-series
representation learning method by combining the advantages of self-supervised
tasks related to contextual, temporal, and transformation consistency. It
allows the network to learn general representations for various downstream
tasks and domains. Specifically, we first adopt data preprocessing to generate
positive and negative pairs for each self-supervised task. The model then
performs contextual, temporal, and transformation contrastive learning and is
optimized jointly using their contrastive losses. We further investigate an
uncertainty weighting approach to enable effective multi-task learning by
considering the contribution of each consistency. We evaluate the proposed
framework on three downstream tasks: time-series classification, forecasting,
and anomaly detection. Experimental results show that our method not only
outperforms the benchmark models on these downstream tasks, but also shows
efficiency in cross-domain transfer learning.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 07:44:06 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Choi",
"Heejeong",
""
],
[
"Kang",
"Pilsung",
""
]
] | TITLE: Multi-Task Self-Supervised Time-Series Representation Learning
ABSTRACT: Time-series representation learning can extract representations from data
with temporal dynamics and sparse labels. When labeled data are sparse but
unlabeled data are abundant, contrastive learning, i.e., a framework to learn a
latent space where similar samples are close to each other while dissimilar
ones are far from each other, has shown outstanding performance. This strategy
can encourage varied consistency of time-series representations depending on
the positive pair selection and contrastive loss. We propose a new time-series
representation learning method by combining the advantages of self-supervised
tasks related to contextual, temporal, and transformation consistency. It
allows the network to learn general representations for various downstream
tasks and domains. Specifically, we first adopt data preprocessing to generate
positive and negative pairs for each self-supervised task. The model then
performs contextual, temporal, and transformation contrastive learning and is
optimized jointly using their contrastive losses. We further investigate an
uncertainty weighting approach to enable effective multi-task learning by
considering the contribution of each consistency. We evaluate the proposed
framework on three downstream tasks: time-series classification, forecasting,
and anomaly detection. Experimental results show that our method not only
outperforms the benchmark models on these downstream tasks, but also shows
efficiency in cross-domain transfer learning.
|
2303.01042 | Yuhu Shang | Yuhu Shang, Xuexiong Luo, Lihong Wang, Hao Peng, Xiankun Zhang, Yimeng
Ren, Kun Liang | Reinforcement Learning Guided Multi-Objective Exam Paper Generation | null | null | null | null | cs.LG cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To reduce the repetitive and complex work of instructors, exam paper
generation (EPG) technique has become a salient topic in the intelligent
education field, which targets at generating high-quality exam paper
automatically according to instructor-specified assessment criteria. The
current advances utilize the ability of heuristic algorithms to optimize
several well-known objective constraints, such as difficulty degree, number of
questions, etc., for producing optimal solutions. However, in real scenarios,
considering other equally relevant objectives (e.g., distribution of exam
scores, skill coverage) is extremely important. Besides, how to develop an
automatic multi-objective solution that finds an optimal subset of questions
from a huge search space of large-sized question datasets and thus composes a
high-quality exam paper is urgent but non-trivial. To this end, we skillfully
design a reinforcement learning guided Multi-Objective Exam Paper Generation
framework, termed MOEPG, to simultaneously optimize three exam domain-specific
objectives including difficulty degree, distribution of exam scores, and skill
coverage. Specifically, to accurately measure the skill proficiency of the
examinee group, we first employ deep knowledge tracing to model the interaction
information between examinees and response logs. We then design the flexible
Exam Q-Network, a function approximator, which automatically selects the
appropriate question to update the exam paper composition process. Later, MOEPG
divides the decision space into multiple subspaces to better guide the updated
direction of the exam paper. Through extensive experiments on two real-world
datasets, we demonstrate that MOEPG is feasible in addressing the multiple
dilemmas of exam paper generation scenario.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 07:55:52 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Shang",
"Yuhu",
""
],
[
"Luo",
"Xuexiong",
""
],
[
"Wang",
"Lihong",
""
],
[
"Peng",
"Hao",
""
],
[
"Zhang",
"Xiankun",
""
],
[
"Ren",
"Yimeng",
""
],
[
"Liang",
"Kun",
""
]
] | TITLE: Reinforcement Learning Guided Multi-Objective Exam Paper Generation
ABSTRACT: To reduce the repetitive and complex work of instructors, exam paper
generation (EPG) technique has become a salient topic in the intelligent
education field, which targets at generating high-quality exam paper
automatically according to instructor-specified assessment criteria. The
current advances utilize the ability of heuristic algorithms to optimize
several well-known objective constraints, such as difficulty degree, number of
questions, etc., for producing optimal solutions. However, in real scenarios,
considering other equally relevant objectives (e.g., distribution of exam
scores, skill coverage) is extremely important. Besides, how to develop an
automatic multi-objective solution that finds an optimal subset of questions
from a huge search space of large-sized question datasets and thus composes a
high-quality exam paper is urgent but non-trivial. To this end, we skillfully
design a reinforcement learning guided Multi-Objective Exam Paper Generation
framework, termed MOEPG, to simultaneously optimize three exam domain-specific
objectives including difficulty degree, distribution of exam scores, and skill
coverage. Specifically, to accurately measure the skill proficiency of the
examinee group, we first employ deep knowledge tracing to model the interaction
information between examinees and response logs. We then design the flexible
Exam Q-Network, a function approximator, which automatically selects the
appropriate question to update the exam paper composition process. Later, MOEPG
divides the decision space into multiple subspaces to better guide the updated
direction of the exam paper. Through extensive experiments on two real-world
datasets, we demonstrate that MOEPG is feasible in addressing the multiple
dilemmas of exam paper generation scenario.
|
2303.01049 | Shuai Xiao | Shuai Xiao, Le Guo, Zaifan Jiang, Lei Lv, Yuanbo Chen, Jun Zhu, Shuang
Yang | Model-based Constrained MDP for Budget Allocation in Sequential
Incentive Marketing | Published at CIKM '19: Proceedings of the 28th ACM International
Conference on Information and Knowledge Management | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Sequential incentive marketing is an important approach for online businesses
to acquire customers, increase loyalty and boost sales. How to effectively
allocate the incentives so as to maximize the return (e.g., business
objectives) under the budget constraint, however, is less studied in the
literature. This problem is technically challenging due to the facts that 1)
the allocation strategy has to be learned using historically logged data, which
is counterfactual in nature, and 2) both the optimality and feasibility (i.e.,
that cost cannot exceed budget) needs to be assessed before being deployed to
online systems. In this paper, we formulate the problem as a constrained Markov
decision process (CMDP). To solve the CMDP problem with logged counterfactual
data, we propose an efficient learning algorithm which combines bisection
search and model-based planning. First, the CMDP is converted into its dual
using Lagrangian relaxation, which is proved to be monotonic with respect to
the dual variable. Furthermore, we show that the dual problem can be solved by
policy learning, with the optimal dual variable being found efficiently via
bisection search (i.e., by taking advantage of the monotonicity). Lastly, we
show that model-based planing can be used to effectively accelerate the joint
optimization process without retraining the policy for every dual variable.
Empirical results on synthetic and real marketing datasets confirm the
effectiveness of our methods.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 08:10:45 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Xiao",
"Shuai",
""
],
[
"Guo",
"Le",
""
],
[
"Jiang",
"Zaifan",
""
],
[
"Lv",
"Lei",
""
],
[
"Chen",
"Yuanbo",
""
],
[
"Zhu",
"Jun",
""
],
[
"Yang",
"Shuang",
""
]
] | TITLE: Model-based Constrained MDP for Budget Allocation in Sequential
Incentive Marketing
ABSTRACT: Sequential incentive marketing is an important approach for online businesses
to acquire customers, increase loyalty and boost sales. How to effectively
allocate the incentives so as to maximize the return (e.g., business
objectives) under the budget constraint, however, is less studied in the
literature. This problem is technically challenging due to the facts that 1)
the allocation strategy has to be learned using historically logged data, which
is counterfactual in nature, and 2) both the optimality and feasibility (i.e.,
that cost cannot exceed budget) needs to be assessed before being deployed to
online systems. In this paper, we formulate the problem as a constrained Markov
decision process (CMDP). To solve the CMDP problem with logged counterfactual
data, we propose an efficient learning algorithm which combines bisection
search and model-based planning. First, the CMDP is converted into its dual
using Lagrangian relaxation, which is proved to be monotonic with respect to
the dual variable. Furthermore, we show that the dual problem can be solved by
policy learning, with the optimal dual variable being found efficiently via
bisection search (i.e., by taking advantage of the monotonicity). Lastly, we
show that model-based planing can be used to effectively accelerate the joint
optimization process without retraining the policy for every dual variable.
Empirical results on synthetic and real marketing datasets confirm the
effectiveness of our methods.
|
2303.01064 | Li Wang | Li Wang, Ying Wah Teh, Mohammed Ali Al-Garadi | Adopting the Multi-answer Questioning Task with an Auxiliary Metric for
Extreme Multi-label Text Classification Utilizing the Label Hierarchy | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Extreme multi-label text classification utilizes the label hierarchy to
partition extreme labels into multiple label groups, turning the task into
simple multi-group multi-label classification tasks. Current research encodes
labels as a vector with fixed length which needs establish multiple classifiers
for different label groups. The problem is how to build only one classifier
without sacrificing the label relationship in the hierarchy. This paper adopts
the multi-answer questioning task for extreme multi-label classification. This
paper also proposes an auxiliary classification evaluation metric. This study
adopts the proposed method and the evaluation metric to the legal domain. The
utilization of legal Berts and the study on task distribution are discussed.
The experiment results show that the proposed hierarchy and multi-answer
questioning task can do extreme multi-label classification for EURLEX dataset.
And in minor/fine-tuning the multi-label classification task, the domain
adapted BERT models could not show apparent advantages in this experiment. The
method is also theoretically applicable to zero-shot learning.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 08:40:31 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Wang",
"Li",
""
],
[
"Teh",
"Ying Wah",
""
],
[
"Al-Garadi",
"Mohammed Ali",
""
]
] | TITLE: Adopting the Multi-answer Questioning Task with an Auxiliary Metric for
Extreme Multi-label Text Classification Utilizing the Label Hierarchy
ABSTRACT: Extreme multi-label text classification utilizes the label hierarchy to
partition extreme labels into multiple label groups, turning the task into
simple multi-group multi-label classification tasks. Current research encodes
labels as a vector with fixed length which needs establish multiple classifiers
for different label groups. The problem is how to build only one classifier
without sacrificing the label relationship in the hierarchy. This paper adopts
the multi-answer questioning task for extreme multi-label classification. This
paper also proposes an auxiliary classification evaluation metric. This study
adopts the proposed method and the evaluation metric to the legal domain. The
utilization of legal Berts and the study on task distribution are discussed.
The experiment results show that the proposed hierarchy and multi-answer
questioning task can do extreme multi-label classification for EURLEX dataset.
And in minor/fine-tuning the multi-label classification task, the domain
adapted BERT models could not show apparent advantages in this experiment. The
method is also theoretically applicable to zero-shot learning.
|
2303.01080 | Xiaoguang Chang | Xiaoguang Chang, Teng Wang, Shaowei Cai and Changyin Sun | LANDMARK: Language-guided Representation Enhancement Framework for Scene
Graph Generation | Revision period in Applied Intelligence (APIN) | null | null | null | cs.CV cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scene graph generation (SGG) is a sophisticated task that suffers from both
complex visual features and dataset long-tail problem. Recently, various
unbiased strategies have been proposed by designing novel loss functions and
data balancing strategies. Unfortunately, these unbiased methods fail to
emphasize language priors in feature refinement perspective. Inspired by the
fact that predicates are highly correlated with semantics hidden in
subject-object pair and global context, we propose LANDMARK (LANguage-guiDed
representationenhanceMent frAmewoRK) that learns predicate-relevant
representations from language-vision interactive patterns, global language
context and pair-predicate correlation. Specifically, we first project object
labels to three distinctive semantic embeddings for different representation
learning. Then, Language Attention Module (LAM) and Experience Estimation
Module (EEM) process subject-object word embeddings to attention vector and
predicate distribution, respectively. Language Context Module (LCM) encodes
global context from each word embed-ding, which avoids isolated learning from
local information. Finally, modules outputs are used to update visual
representations and SGG model's prediction. All language representations are
purely generated from object categories so that no extra knowledge is needed.
This framework is model-agnostic and consistently improves performance on
existing SGG models. Besides, representation-level unbiased strategies endow
LANDMARK the advantage of compatibility with other methods. Code is available
at https://github.com/rafa-cxg/PySGG-cxg.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 09:03:11 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Chang",
"Xiaoguang",
""
],
[
"Wang",
"Teng",
""
],
[
"Cai",
"Shaowei",
""
],
[
"Sun",
"Changyin",
""
]
] | TITLE: LANDMARK: Language-guided Representation Enhancement Framework for Scene
Graph Generation
ABSTRACT: Scene graph generation (SGG) is a sophisticated task that suffers from both
complex visual features and dataset long-tail problem. Recently, various
unbiased strategies have been proposed by designing novel loss functions and
data balancing strategies. Unfortunately, these unbiased methods fail to
emphasize language priors in feature refinement perspective. Inspired by the
fact that predicates are highly correlated with semantics hidden in
subject-object pair and global context, we propose LANDMARK (LANguage-guiDed
representationenhanceMent frAmewoRK) that learns predicate-relevant
representations from language-vision interactive patterns, global language
context and pair-predicate correlation. Specifically, we first project object
labels to three distinctive semantic embeddings for different representation
learning. Then, Language Attention Module (LAM) and Experience Estimation
Module (EEM) process subject-object word embeddings to attention vector and
predicate distribution, respectively. Language Context Module (LCM) encodes
global context from each word embed-ding, which avoids isolated learning from
local information. Finally, modules outputs are used to update visual
representations and SGG model's prediction. All language representations are
purely generated from object categories so that no extra knowledge is needed.
This framework is model-agnostic and consistently improves performance on
existing SGG models. Besides, representation-level unbiased strategies endow
LANDMARK the advantage of compatibility with other methods. Code is available
at https://github.com/rafa-cxg/PySGG-cxg.
|
2303.01086 | Chunfeng Wang | Chunfeng Wang, Peisong Huang, Yuxiang Zou, Haoyu Zhang, Shichao Liu,
Xiang Yin, Zejun Ma | LiteG2P: A fast, light and high accuracy model for grapheme-to-phoneme
conversion | Accepted by ICASSP2023 | null | null | null | cs.CL cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As a key component of automated speech recognition (ASR) and the front-end in
text-to-speech (TTS), grapheme-to-phoneme (G2P) plays the role of converting
letters to their corresponding pronunciations. Existing methods are either slow
or poor in performance, and are limited in application scenarios, particularly
in the process of on-device inference. In this paper, we integrate the
advantages of both expert knowledge and connectionist temporal classification
(CTC) based neural network and propose a novel method named LiteG2P which is
fast, light and theoretically parallel. With the carefully leading design,
LiteG2P can be applied both on cloud and on device. Experimental results on the
CMU dataset show that the performance of the proposed method is superior to the
state-of-the-art CTC based method with 10 times fewer parameters, and even
comparable to the state-of-the-art Transformer-based sequence-to-sequence model
with less parameters and 33 times less computation.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 09:16:21 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Wang",
"Chunfeng",
""
],
[
"Huang",
"Peisong",
""
],
[
"Zou",
"Yuxiang",
""
],
[
"Zhang",
"Haoyu",
""
],
[
"Liu",
"Shichao",
""
],
[
"Yin",
"Xiang",
""
],
[
"Ma",
"Zejun",
""
]
] | TITLE: LiteG2P: A fast, light and high accuracy model for grapheme-to-phoneme
conversion
ABSTRACT: As a key component of automated speech recognition (ASR) and the front-end in
text-to-speech (TTS), grapheme-to-phoneme (G2P) plays the role of converting
letters to their corresponding pronunciations. Existing methods are either slow
or poor in performance, and are limited in application scenarios, particularly
in the process of on-device inference. In this paper, we integrate the
advantages of both expert knowledge and connectionist temporal classification
(CTC) based neural network and propose a novel method named LiteG2P which is
fast, light and theoretically parallel. With the carefully leading design,
LiteG2P can be applied both on cloud and on device. Experimental results on the
CMU dataset show that the performance of the proposed method is superior to the
state-of-the-art CTC based method with 10 times fewer parameters, and even
comparable to the state-of-the-art Transformer-based sequence-to-sequence model
with less parameters and 33 times less computation.
|
2303.01094 | Congchi Yin | Congchi Yin, Piji Li and Zhaochun Ren | CTRLStruct: Dialogue Structure Learning for Open-Domain Response
Generation | 12 pages, to be published in The Web Conference 2023 | null | 10.1145/3543507.3583285 | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Dialogue structure discovery is essential in dialogue generation.
Well-structured topic flow can leverage background information and predict
future topics to help generate controllable and explainable responses. However,
most previous work focused on dialogue structure learning in task-oriented
dialogue other than open-domain dialogue which is more complicated and
challenging. In this paper, we present a new framework CTRLStruct for dialogue
structure learning to effectively explore topic-level dialogue clusters as well
as their transitions with unlabelled information. Precisely, dialogue
utterances encoded by bi-directional Transformer are further trained through a
special designed contrastive learning task to improve representation. Then we
perform clustering to utterance-level representations and form topic-level
clusters that can be considered as vertices in dialogue structure graph. The
edges in the graph indicating transition probability between vertices are
calculated by mimicking expert behavior in datasets. Finally, dialogue
structure graph is integrated into dialogue model to perform controlled
response generation. Experiments on two popular open-domain dialogue datasets
show our model can generate more coherent responses compared to some excellent
dialogue models, as well as outperform some typical sentence embedding methods
in dialogue utterance representation. Code is available in GitHub.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 09:27:11 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Yin",
"Congchi",
""
],
[
"Li",
"Piji",
""
],
[
"Ren",
"Zhaochun",
""
]
] | TITLE: CTRLStruct: Dialogue Structure Learning for Open-Domain Response
Generation
ABSTRACT: Dialogue structure discovery is essential in dialogue generation.
Well-structured topic flow can leverage background information and predict
future topics to help generate controllable and explainable responses. However,
most previous work focused on dialogue structure learning in task-oriented
dialogue other than open-domain dialogue which is more complicated and
challenging. In this paper, we present a new framework CTRLStruct for dialogue
structure learning to effectively explore topic-level dialogue clusters as well
as their transitions with unlabelled information. Precisely, dialogue
utterances encoded by bi-directional Transformer are further trained through a
special designed contrastive learning task to improve representation. Then we
perform clustering to utterance-level representations and form topic-level
clusters that can be considered as vertices in dialogue structure graph. The
edges in the graph indicating transition probability between vertices are
calculated by mimicking expert behavior in datasets. Finally, dialogue
structure graph is integrated into dialogue model to perform controlled
response generation. Experiments on two popular open-domain dialogue datasets
show our model can generate more coherent responses compared to some excellent
dialogue models, as well as outperform some typical sentence embedding methods
in dialogue utterance representation. Code is available in GitHub.
|
2303.01099 | Adrian Galdran | Adrian Galdran, Johan Verjans, Gustavo Carneiro, Miguel A. Gonz\'alez
Ballester | Multi-Head Multi-Loss Model Calibration | Under review | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Delivering meaningful uncertainty estimates is essential for a successful
deployment of machine learning models in the clinical practice. A central
aspect of uncertainty quantification is the ability of a model to return
predictions that are well-aligned with the actual probability of the model
being correct, also known as model calibration. Although many methods have been
proposed to improve calibration, no technique can match the simple, but
expensive approach of training an ensemble of deep neural networks. In this
paper we introduce a form of simplified ensembling that bypasses the costly
training and inference of deep ensembles, yet it keeps its calibration
capabilities. The idea is to replace the common linear classifier at the end of
a network by a set of heads that are supervised with different loss functions
to enforce diversity on their predictions. Specifically, each head is trained
to minimize a weighted Cross-Entropy loss, but the weights are different among
the different branches. We show that the resulting averaged predictions can
achieve excellent calibration without sacrificing accuracy in two challenging
datasets for histopathological and endoscopic image classification. Our
experiments indicate that Multi-Head Multi-Loss classifiers are inherently
well-calibrated, outperforming other recent calibration techniques and even
challenging Deep Ensembles' performance. Code to reproduce our experiments can
be found at \url{https://github.com/agaldran/mhml_calibration} .
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 09:32:32 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Galdran",
"Adrian",
""
],
[
"Verjans",
"Johan",
""
],
[
"Carneiro",
"Gustavo",
""
],
[
"Ballester",
"Miguel A. González",
""
]
] | TITLE: Multi-Head Multi-Loss Model Calibration
ABSTRACT: Delivering meaningful uncertainty estimates is essential for a successful
deployment of machine learning models in the clinical practice. A central
aspect of uncertainty quantification is the ability of a model to return
predictions that are well-aligned with the actual probability of the model
being correct, also known as model calibration. Although many methods have been
proposed to improve calibration, no technique can match the simple, but
expensive approach of training an ensemble of deep neural networks. In this
paper we introduce a form of simplified ensembling that bypasses the costly
training and inference of deep ensembles, yet it keeps its calibration
capabilities. The idea is to replace the common linear classifier at the end of
a network by a set of heads that are supervised with different loss functions
to enforce diversity on their predictions. Specifically, each head is trained
to minimize a weighted Cross-Entropy loss, but the weights are different among
the different branches. We show that the resulting averaged predictions can
achieve excellent calibration without sacrificing accuracy in two challenging
datasets for histopathological and endoscopic image classification. Our
experiments indicate that Multi-Head Multi-Loss classifiers are inherently
well-calibrated, outperforming other recent calibration techniques and even
challenging Deep Ensembles' performance. Code to reproduce our experiments can
be found at \url{https://github.com/agaldran/mhml_calibration} .
|
2303.01112 | Sora Takashima | Sora Takashima, Ryo Hayamizu, Nakamasa Inoue, Hirokatsu Kataoka, Rio
Yokota | Visual Atoms: Pre-training Vision Transformers with Sinusoidal Waves | Accepted to CVPR 2023 | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Formula-driven supervised learning (FDSL) has been shown to be an effective
method for pre-training vision transformers, where ExFractalDB-21k was shown to
exceed the pre-training effect of ImageNet-21k. These studies also indicate
that contours mattered more than textures when pre-training vision
transformers. However, the lack of a systematic investigation as to why these
contour-oriented synthetic datasets can achieve the same accuracy as real
datasets leaves much room for skepticism. In the present work, we develop a
novel methodology based on circular harmonics for systematically investigating
the design space of contour-oriented synthetic datasets. This allows us to
efficiently search the optimal range of FDSL parameters and maximize the
variety of synthetic images in the dataset, which we found to be a critical
factor. When the resulting new dataset VisualAtom-21k is used for pre-training
ViT-Base, the top-1 accuracy reached 83.7% when fine-tuning on ImageNet-1k.
This is close to the top-1 accuracy (84.2%) achieved by JFT-300M pre-training,
while the number of images is 1/14. Unlike JFT-300M which is a static dataset,
the quality of synthetic datasets will continue to improve, and the current
work is a testament to this possibility. FDSL is also free of the common issues
associated with real images, e.g. privacy/copyright issues, labeling
costs/errors, and ethical biases.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 09:47:28 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Takashima",
"Sora",
""
],
[
"Hayamizu",
"Ryo",
""
],
[
"Inoue",
"Nakamasa",
""
],
[
"Kataoka",
"Hirokatsu",
""
],
[
"Yokota",
"Rio",
""
]
] | TITLE: Visual Atoms: Pre-training Vision Transformers with Sinusoidal Waves
ABSTRACT: Formula-driven supervised learning (FDSL) has been shown to be an effective
method for pre-training vision transformers, where ExFractalDB-21k was shown to
exceed the pre-training effect of ImageNet-21k. These studies also indicate
that contours mattered more than textures when pre-training vision
transformers. However, the lack of a systematic investigation as to why these
contour-oriented synthetic datasets can achieve the same accuracy as real
datasets leaves much room for skepticism. In the present work, we develop a
novel methodology based on circular harmonics for systematically investigating
the design space of contour-oriented synthetic datasets. This allows us to
efficiently search the optimal range of FDSL parameters and maximize the
variety of synthetic images in the dataset, which we found to be a critical
factor. When the resulting new dataset VisualAtom-21k is used for pre-training
ViT-Base, the top-1 accuracy reached 83.7% when fine-tuning on ImageNet-1k.
This is close to the top-1 accuracy (84.2%) achieved by JFT-300M pre-training,
while the number of images is 1/14. Unlike JFT-300M which is a static dataset,
the quality of synthetic datasets will continue to improve, and the current
work is a testament to this possibility. FDSL is also free of the common issues
associated with real images, e.g. privacy/copyright issues, labeling
costs/errors, and ethical biases.
|
2303.01156 | Max Hahn-Klimroth | Yannick Gerstorfer, Lena Krieg, Max Hahn-Klimroth | A Notion of Feature Importance by Decorrelation and Detection of Trends
by Random Forest Regression | null | null | null | null | stat.ML cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In many studies, we want to determine the influence of certain features on a
dependent variable. More specifically, we are interested in the strength of the
influence -- i.e., is the feature relevant? -- and, if so, how the feature
influences the dependent variable. Recently, data-driven approaches such as
\emph{random forest regression} have found their way into applications
(Boulesteix et al., 2012). These models allow to directly derive measures of
feature importance, which are a natural indicator of the strength of the
influence. For the relevant features, the correlation or rank correlation
between the feature and the dependent variable has typically been used to
determine the nature of the influence. More recent methods, some of which can
also measure interactions between features, are based on a modeling approach.
In particular, when machine learning models are used, SHAP scores are a recent
and prominent method to determine these trends (Lundberg et al., 2017).
In this paper, we introduce a novel notion of feature importance based on the
well-studied Gram-Schmidt decorrelation method. Furthermore, we propose two
estimators for identifying trends in the data using random forest regression,
the so-called absolute and relative transversal rate. We empirically compare
the properties of our estimators with those of well-established estimators on a
variety of synthetic and real-world datasets.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 11:01:49 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Gerstorfer",
"Yannick",
""
],
[
"Krieg",
"Lena",
""
],
[
"Hahn-Klimroth",
"Max",
""
]
] | TITLE: A Notion of Feature Importance by Decorrelation and Detection of Trends
by Random Forest Regression
ABSTRACT: In many studies, we want to determine the influence of certain features on a
dependent variable. More specifically, we are interested in the strength of the
influence -- i.e., is the feature relevant? -- and, if so, how the feature
influences the dependent variable. Recently, data-driven approaches such as
\emph{random forest regression} have found their way into applications
(Boulesteix et al., 2012). These models allow to directly derive measures of
feature importance, which are a natural indicator of the strength of the
influence. For the relevant features, the correlation or rank correlation
between the feature and the dependent variable has typically been used to
determine the nature of the influence. More recent methods, some of which can
also measure interactions between features, are based on a modeling approach.
In particular, when machine learning models are used, SHAP scores are a recent
and prominent method to determine these trends (Lundberg et al., 2017).
In this paper, we introduce a novel notion of feature importance based on the
well-studied Gram-Schmidt decorrelation method. Furthermore, we propose two
estimators for identifying trends in the data using random forest regression,
the so-called absolute and relative transversal rate. We empirically compare
the properties of our estimators with those of well-established estimators on a
variety of synthetic and real-world datasets.
|
2303.01158 | Matthias Cosler | Matthias Cosler, Frederik Schmitt, Christopher Hahn, Bernd Finkbeiner | Iterative Circuit Repair Against Formal Specifications | To appear at ICLR'23 | null | null | null | cs.LG cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a deep learning approach for repairing sequential circuits against
formal specifications given in linear-time temporal logic (LTL). Given a
defective circuit and its formal specification, we train Transformer models to
output circuits that satisfy the corresponding specification. We propose a
separated hierarchical Transformer for multimodal representation learning of
the formal specification and the circuit. We introduce a data generation
algorithm that enables generalization to more complex specifications and
out-of-distribution datasets. In addition, our proposed repair mechanism
significantly improves the automated synthesis of circuits from LTL
specifications with Transformers. It improves the state-of-the-art by $6.8$
percentage points on held-out instances and $11.8$ percentage points on an
out-of-distribution dataset from the annual reactive synthesis competition.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 11:05:10 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Cosler",
"Matthias",
""
],
[
"Schmitt",
"Frederik",
""
],
[
"Hahn",
"Christopher",
""
],
[
"Finkbeiner",
"Bernd",
""
]
] | TITLE: Iterative Circuit Repair Against Formal Specifications
ABSTRACT: We present a deep learning approach for repairing sequential circuits against
formal specifications given in linear-time temporal logic (LTL). Given a
defective circuit and its formal specification, we train Transformer models to
output circuits that satisfy the corresponding specification. We propose a
separated hierarchical Transformer for multimodal representation learning of
the formal specification and the circuit. We introduce a data generation
algorithm that enables generalization to more complex specifications and
out-of-distribution datasets. In addition, our proposed repair mechanism
significantly improves the automated synthesis of circuits from LTL
specifications with Transformers. It improves the state-of-the-art by $6.8$
percentage points on held-out instances and $11.8$ percentage points on an
out-of-distribution dataset from the annual reactive synthesis competition.
|
2303.01166 | Zhixing Hou | Zhixing Hou, Yuzhang Shang, Tian Gao, Yan Yan | BPT: Binary Point Cloud Transformer for Place Recognition | Submitted to the IEEE/RSJ International Conference on Intelligent
Robots (IROS 2023) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Place recognition, an algorithm to recognize the re-visited places, plays the
role of back-end optimization trigger in a full SLAM system. Many works
equipped with deep learning tools, such as MLP, CNN, and transformer, have
achieved great improvements in this research field. Point cloud transformer is
one of the excellent frameworks for place recognition applied in robotics, but
with large memory consumption and expensive computation, it is adverse to
widely deploy the various point cloud transformer networks in mobile or
embedded devices. To solve this issue, we propose a binary point cloud
transformer for place recognition. As a result, a 32-bit full-precision model
can be reduced to a 1-bit model with less memory occupation and faster
binarized bitwise operations. To our best knowledge, this is the first binary
point cloud transformer that can be deployed on mobile devices for online
applications such as place recognition. Experiments on several standard
benchmarks demonstrate that the proposed method can get comparable results with
the corresponding full-precision transformer model and even outperform some
full-precision deep learning methods. For example, the proposed method achieves
93.28% at the top @1% and 85.74% at the top @1% on the Oxford RobotCar dataset
in terms of the metric of the average recall rate. Meanwhile, the size and
floating point operations of the model with the same transformer structure
reduce 56.1% and 34.1% respectively from original precision to binary
precision.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 11:15:59 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Hou",
"Zhixing",
""
],
[
"Shang",
"Yuzhang",
""
],
[
"Gao",
"Tian",
""
],
[
"Yan",
"Yan",
""
]
] | TITLE: BPT: Binary Point Cloud Transformer for Place Recognition
ABSTRACT: Place recognition, an algorithm to recognize the re-visited places, plays the
role of back-end optimization trigger in a full SLAM system. Many works
equipped with deep learning tools, such as MLP, CNN, and transformer, have
achieved great improvements in this research field. Point cloud transformer is
one of the excellent frameworks for place recognition applied in robotics, but
with large memory consumption and expensive computation, it is adverse to
widely deploy the various point cloud transformer networks in mobile or
embedded devices. To solve this issue, we propose a binary point cloud
transformer for place recognition. As a result, a 32-bit full-precision model
can be reduced to a 1-bit model with less memory occupation and faster
binarized bitwise operations. To our best knowledge, this is the first binary
point cloud transformer that can be deployed on mobile devices for online
applications such as place recognition. Experiments on several standard
benchmarks demonstrate that the proposed method can get comparable results with
the corresponding full-precision transformer model and even outperform some
full-precision deep learning methods. For example, the proposed method achieves
93.28% at the top @1% and 85.74% at the top @1% on the Oxford RobotCar dataset
in terms of the metric of the average recall rate. Meanwhile, the size and
floating point operations of the model with the same transformer structure
reduce 56.1% and 34.1% respectively from original precision to binary
precision.
|
2303.01178 | Manuel Cossio | Manuel Cossio | Augmenting Medical Imaging: A Comprehensive Catalogue of 65 Techniques
for Enhanced Data Analysis | 12 pages | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | In the realm of medical imaging, the training of machine learning models
necessitates a large and varied training dataset to ensure robustness and
interoperability. However, acquiring such diverse and heterogeneous data can be
difficult due to the need for expert labeling of each image and privacy
concerns associated with medical data. To circumvent these challenges, data
augmentation has emerged as a promising and cost-effective technique for
increasing the size and diversity of the training dataset. In this study, we
provide a comprehensive review of the specific data augmentation techniques
employed in medical imaging and explore their benefits. We conducted an
in-depth study of all data augmentation techniques used in medical imaging,
identifying 11 different purposes and collecting 65 distinct techniques. The
techniques were operationalized into spatial transformation-based, color and
contrast adjustment-based, noise-based, deformation-based, data mixing-based,
filters and mask-based, division-based, multi-scale and multi-view-based, and
meta-learning-based categories. We observed that some techniques require manual
specification of all parameters, while others rely on automation to adjust the
type and magnitude of augmentation based on task requirements. The utilization
of these techniques enables the development of more robust models that can be
applied in domains with limited or challenging data availability. It is
expected that the list of available techniques will expand in the future,
providing researchers with additional options to consider.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 11:47:55 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Cossio",
"Manuel",
""
]
] | TITLE: Augmenting Medical Imaging: A Comprehensive Catalogue of 65 Techniques
for Enhanced Data Analysis
ABSTRACT: In the realm of medical imaging, the training of machine learning models
necessitates a large and varied training dataset to ensure robustness and
interoperability. However, acquiring such diverse and heterogeneous data can be
difficult due to the need for expert labeling of each image and privacy
concerns associated with medical data. To circumvent these challenges, data
augmentation has emerged as a promising and cost-effective technique for
increasing the size and diversity of the training dataset. In this study, we
provide a comprehensive review of the specific data augmentation techniques
employed in medical imaging and explore their benefits. We conducted an
in-depth study of all data augmentation techniques used in medical imaging,
identifying 11 different purposes and collecting 65 distinct techniques. The
techniques were operationalized into spatial transformation-based, color and
contrast adjustment-based, noise-based, deformation-based, data mixing-based,
filters and mask-based, division-based, multi-scale and multi-view-based, and
meta-learning-based categories. We observed that some techniques require manual
specification of all parameters, while others rely on automation to adjust the
type and magnitude of augmentation based on task requirements. The utilization
of these techniques enables the development of more robust models that can be
applied in domains with limited or challenging data availability. It is
expected that the list of available techniques will expand in the future,
providing researchers with additional options to consider.
|
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.