Update app.py

app.py

@@ -27,20 +27,41 @@ def process_video(video, sitting, hands, location, screen):
 def process_and_display_json(video_description): 
     json_response = process_description(video_description)
     return json_response
-    
-    # --- Gradio Interface --- 
-video = gr.Video(label="Video")
-sitting = gr.Checkbox(label="Sitting/Standing")
-hands = gr.Checkbox(label="Hands Free/Not Free")
-location = gr.Checkbox(label="Indoors/Outdoors")
-screen = gr.Checkbox(label="Screen Interaction")
 
-video_description = gr.Textbox(label="Video Description", show_label=True, show_copy_button=True)
-json_output = gr.JSON(label="JSON Output") 
 
+# --- Function to handle JSON processing ---
+def process_and_display_json(video_description):
+    json_response = process_description(video_description)
+    return json_response
+
+# --- Gradio Blocks Interface --- 
+with gr.Blocks(title="GSoC Super Raid Annotator", theme=gr.themes.Soft(primary_hue="red", secondary_hue="red")) as demo:
+    gr.Markdown("Annotate Videos")
+    gr.Markdown(
+        "<p style='text-align: center'><a href='https://github.com/OpenBMB/MiniCPM-V' target='_blank'>Model GitHub Repo</a> | <a href='https://huggingface.co/openbmb/MiniCPM-V-2_6' target='_blank'>Model Page</a></p>"
+    )
+
+    with gr.Row():
+        with gr.Column():
+            # Input components
+            video = gr.Video(label="Video")
+            sitting = gr.Checkbox(label="Sitting/Standing")
+            hands = gr.Checkbox(label="Hands Free/Not Free")
+            location = gr.Checkbox(label="Indoors/Outdoors")
+            screen = gr.Checkbox(label="Screen Interaction")
 
-# Examples for the interface
-examples = [
+            # Submit buttons
+            with gr.Row():
+                process_video_btn = gr.Button("Process Video")
+                process_json_btn = gr.Button("Process JSON")
+
+        with gr.Column():
+            # Output components
+            video_description = gr.Textbox(label="Video Description", show_label=True, show_copy_button=True)
+            json_output = gr.JSON(label="JSON Output")
+
+    # Examples for the interface
+    examples = [
     ["videos/2016-01-01_0100_US_KNBC_Channel_4_News_1867.16-1871.38_now.mp4",],
     ["videos/2016-01-01_0200_US_KNBC_Channel_4_News_1329.12-1333.29_tonight.mp4",],
     ["videos/2016-01-01_0830_US_KNBC_Tonight_Show_with_Jimmy_Fallon_725.45-729.76_tonight.mp4",],
@@ -52,40 +73,20 @@ examples = [
     ["videos/2016-01-01_1300_US_KNBC_Today_in_LA_at_5am_12.46-16.95_this_morning.mp4"],
     ["videos/2016-01-05_0200_US_KNBC_Channel_4_News_1561.29-1565.95_next_week.mp4"],
     ["videos/2016-01-28_0700_US_KNBC_Channel_4_News_at_11PM_629.56-633.99_in_the_future.mp4"]
-]
-
-# Title, description, and article for the interface
-title = "GSoC Super Raid Annotator"
-description = "Annotate Videos"
-article = "<p style='text-align: center'><a href='https://github.com/OpenBMB/MiniCPM-V' target='_blank'>Model GitHub Repo</a> | <a href='https://huggingface.co/openbmb/MiniCPM-V-2_6' target='_blank'>Model Page</a></p>"
-
-custom_theme = gr.themes.Soft(
-    primary_hue="blue", 
-    secondary_hue="blue",
-    neutral_hue="zinc"
-)
+    ]
 
-# --- Create the Interface --- 
-interface = gr.Interface(
-    fn=process_video,
-    inputs=[video, sitting, hands, location, screen],
-    outputs=video_description, 
-    examples=examples,
-    title=title,
-    description=description,
-    article=article,
-    theme=custom_theme,
-    allow_flagging="never",
-    live=True # Add live=True here
-)
+    # Event handling
+    process_video_btn.click(
+        fn=process_video,
+        inputs=[video, sitting, hands, location, screen],
+        outputs=video_description,
+    )
 
-# --- Add the JSON processing component with live updates ---
-interface.add_component(
-    fn=process_and_display_json,
-    inputs=video_description,  # This will take output from process_video
-    outputs=json_output,
-    live=True,               
-    live_inputs=[0]          # Update based on the first input (video_description)
-)
+    process_json_btn.click(
+        fn=process_and_display_json,
+        inputs=video_description,
+        outputs=json_output,
+    )
 
-interface.launch(debug=False)
+# Launch the interface
+demo.launch(debug=False)