Update app.py

app.py

@@ -5,7 +5,7 @@ client = Client("https://vikhyatk-moondream1.hf.space/")
 import cv2
 from moviepy.editor import *
 
-# 1. extract and store 1 image every 5 images from video input
+# 1. extract and store 1 image every 24 images from video input
 # 2. extract audio
 # 3. for each image from extracted_images, get caption from caption model and concatenate into list
 # 4. for audio, ask audio questioning model to describe sound/scene
@@ -114,10 +114,14 @@ def process_image(image_in):
     '''
 
 def extract_audio(video_path):
-    video_clip = VideoFileClip(video_path)
-    audio_clip = video_clip.audio
-    audio_clip.write_audiofile("output_audio.mp3")
-    return "output_audio.mp3"
+    # Check if the video has audio
+    if video_clip.audio is not None:
+        audio_clip = video_clip.audio
+        audio_clip.write_audiofile("output_audio.mp3")
+        return "output_audio.mp3"
+    else:
+        print("The video does not have any audio.")
+        return None
 
 def get_salmonn(audio_in):
     salmonn_prompt = "Please describe the audio"
@@ -167,10 +171,14 @@ def infer(video_in):
 
     # Extract audio from video
     extracted_audio = extract_audio(video_in)
-    print(extracted_audio)
 
-    # Get description of audio content
-    audio_content_described = get_salmonn(extracted_audio)
+    if extracted is not None :
+        print(extracted_audio)
+
+        # Get description of audio content
+        audio_content_described = get_salmonn(extracted_audio)
+    else : 
+        audio_content_described = "Video has no sound."
 
     # Assemble captions
     formatted_captions = f"""
@@ -199,10 +207,13 @@ with gr.Blocks(css=css) as demo :
         gr.HTML("""
         <h2 style="text-align: center;">Soft video understanding</h2>
         """)
-        video_in = gr.Video(label="Video input")
-        video_cut = gr.Video(label="Video cut")
-        submit_btn = gr.Button("Submit")
-        video_description = gr.Textbox(label="Video description", elem_id="video-text")
+        with gr.Row():
+            video_in = gr.Video(label="Video input")
+            with gr.Column():
+                video_cut = gr.Video(label="Video cut to 10 seconds")
+                submit_btn = gr.Button("Submit")
+                video_description = gr.Textbox(label="Video description", elem_id="video-text")
+    
     video_in.upload(
         fn = trim_video,
         inputs = [video_in],