app.py

import streamlit as st  # Streamlit for building the web application
from transformers import pipeline  # Hugging Face Transformers pipeline for models
from PIL import Image  # PIL for handling image files

# Function to convert image to text
def img2text(image):
    # Load the image-to-text model
    image_to_text_model = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
    # Generate a caption for the image
    text = image_to_text_model(image)[0]["generated_text"]
    return text  # Return the generated caption

# Function to generate a story based on the caption
def text2story(text):
    # Load the text generation model
    story_model = pipeline("text-generation", model="gpt2")
    # Generate a story based on the input text
    story_text = story_model(f"Once upon a time, {text}.", max_length=100, num_return_sequences=1, do_sample=True, top_k=50)
    return story_text[0]["generated_text"]  # Return the generated story

# Function to convert text to audio
def text2audio(story_text):
    # Load the text-to-speech model
    text_to_audio_model = pipeline("text-to-speech", model="facebook/mms-tts-eng")
    # Generate audio data from the story text
    audio_data = text_to_audio_model(story_text)
    return audio_data  # Return the audio data

# Main part of the application
st.set_page_config(page_title="Your Image to Audio Story", page_icon="🦜")  # Set the title and icon of the app
st.header("Storytelling From Your Image")  # Header for the application
uploaded_file = st.file_uploader("Select an Image...", type=["jpg", "jpeg", "png"])  # File uploader for images

if uploaded_file is not None:
    # Open and read the uploaded image
    image = Image.open(uploaded_file)  # Use PIL to open the uploaded image
    st.image(image, caption="Uploaded Image", use_container_width=True)  # Display the uploaded image

    # Stage 1: Image to Text
    st.text('Processing image to text...')  # Inform the user about the processing stage
    scenario = img2text(image)  # Get the caption for the uploaded image
    st.write("Caption:", scenario)  # Display the generated caption

    # Stage 2: Text to Story
    st.text('Generating a story...')  # Inform the user about the story generation stage
    story = text2story(scenario)  # Generate a story based on the caption
    st.write("Story:", story)  # Display the generated story

    # Stage 3: Story to Audio data
    st.text('Generating audio data...')  # Inform the user about the audio generation stage
    audio_data = text2audio(story)  # Convert the generated story into audio

    # Play button for the audio
    if st.button("Play Audio"):  # Create a button to play the audio
        st.audio(audio_data['audio'], 
                 format="audio/wav", 
                 start_time=0, 
                 sample_rate=audio_data['sampling_rate'])  # Play the audio