Update appImage.py

appImage.py

@@ -49,7 +49,8 @@ from PIL import Image
 import tempfile
 import os
 import torch
-
+from gtts import gTTS
+import uuid
 # Load model
 try:
     processor = AutoProcessor.from_pretrained("microsoft/git-large-coco")
@@ -84,14 +85,26 @@ async def caption_image(file: UploadFile):
             contents = await file.read()
             tmp.write(contents)
             tmp_path = tmp.name
-
+        
         # Generate caption
         caption = generate_caption(tmp_path)
+        os.remove(tmp_path)
 
         # Handle errors inside generate_caption
         if caption.startswith("Error"):
             return {"error": caption}
-        return {"caption": caption}
+
+        # Now generate TTS audio for the caption
+        tts = gTTS(text=caption, lang="en")
+        audio_filename = f"{uuid.uuid4()}.mp3"
+        audio_path = os.path.join(tempfile.gettempdir(), audio_filename)
+        tts.save(audio_path)
+
+        # Return both caption and audio URL
+        return {
+            "caption": caption,
+            "audio": f"/files/{audio_filename}"
+        }
 
     except Exception as e:
         return {"error": f"Failed to generate caption: {str(e)}"}