Update app.py

app.py

@@ -2,7 +2,6 @@ import gradio as gr
 from gradio_client import Client
 import os
 import tempfile
-from PIL import Image
 
 BACKEND_URL = os.environ.get("BACKEND_URL", "").strip()
 
@@ -17,153 +16,108 @@ except Exception as e:
     print(f"Backend not available: {e}")
 
 def process_media(file_obj, webcam_img, model_type, conf_thresh, max_dets, task_type):
-    """Main processing function"""
+    """Main processing function - expects 5 outputs"""
     if not client:
         return [gr.update()] * 5
     
     try:
-        # Handle webcam image - save PIL to temp file
+        # Handle webcam image - need to pass as file path
         webcam_file = None
         if webcam_img is not None:
             with tempfile.NamedTemporaryFile(suffix='.png', delete=False) as tmp:
                 webcam_img.save(tmp, 'PNG')
                 webcam_file = tmp.name
         
-        # Call backend - process_media is likely at index 3
+        # Call backend
         result = client.predict(
             file_obj,
-            webcam_file,  # Pass file path instead of PIL object
+            webcam_file,
             model_type,
             conf_thresh,
             max_dets,
             task_type,
-            fn_index=3  # process_media endpoint
+            fn_index=3
         )
         
-        # Clean up temp file
+        # Clean up
         if webcam_file and os.path.exists(webcam_file):
             os.unlink(webcam_file)
-            
-        return result
         
+        # Backend returns 7 values but we only need the last 5
+        # Skip the first 2 (input_file and input_webcam updates)
+        if len(result) == 7:
+            return result[2:]  # Return only the display components
+        else:
+            return result[:5]  # Safety fallback
+            
     except Exception as e:
         print(f"Error in process_media: {e}")
         return [gr.update()] * 5
 
-def clear_all():
-    """Clear all inputs and outputs"""
-    try:
-        # Call backend clear function at index 4
-        return client.predict(fn_index=4)
-    except:
-        # Fallback to local clear
-        return [gr.update(value=None)] * 7
-
-# Build the interface
+# Build simplified interface
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("<center><h1>PrimateFace Detection, Pose Estimation, and Gaze Estimation Demo</h1></center>")
+    gr.Markdown("# 🐵 PrimateFace Detection, Pose Estimation, and Gaze Demo")
     
     if not backend_available:
-        gr.Markdown("### 🔴 GPU Server Offline - Please check back later")
+        gr.Markdown("### 🔴 GPU Server Offline")
     else:
-        gr.Markdown("Upload an image/video or use your webcam. Click 'Detect Faces' for results.")
-        
         with gr.Row():
-            with gr.Column(scale=1):
+            with gr.Column():
                 with gr.Tabs():
-                    with gr.TabItem("Upload File"):
-                        input_file = gr.File(label="Upload Image or Video Here", file_types=["image", "video"])
-                        display_raw_image_file = gr.Image(label="Raw Image Preview", type="pil", visible=False)
-                        display_raw_video_file = gr.Video(label="Raw Video Preview", visible=False)
+                    with gr.TabItem("Upload"):
+                        input_file = gr.File(label="Upload Image/Video", file_types=["image", "video"])
+                        display_raw_image_file = gr.Image(visible=False)
+                        display_raw_video_file = gr.Video(visible=False)
                     
                     with gr.TabItem("Webcam"):
-                        gr.Markdown("Click on feed or press Enter to capture")
                         input_webcam = gr.Image(sources=["webcam"], type="pil")
-                        display_raw_image_webcam = gr.Image(label="Captured Snapshot", type="pil", visible=False)
+                        display_raw_image_webcam = gr.Image(visible=False)
                 
-                clear_all_button = gr.Button("Clear All Inputs & Outputs")
+                clear_button = gr.Button("Clear All")
             
-            with gr.Column(scale=1):
-                gr.Markdown("### Processed Output")
-                display_processed_image = gr.Image(label="Processed Image", type="pil", visible=False)
-                display_processed_video = gr.Video(label="Processed Video", visible=False)
-        
-        # Examples with preview
-        example_images = [
-            "images/allocebus_000003.jpeg",
-            "images/tarsius_000120.jpeg",
-            "images/nasalis_proboscis-monkey.png",
-            "images/macaca_000032.jpeg",
-            "images/mandrillus_000011.jpeg",
-            "images/pongo_000006.jpeg"
-        ]
+            with gr.Column():
+                gr.Markdown("### Output")
+                display_processed_image = gr.Image(visible=False)
+                display_processed_video = gr.Video(visible=False)
         
-        with gr.Row():
-            gr.Examples(
-                examples=[[img] for img in example_images],
-                inputs=input_file,
-                label="Example Images (Click to load)",
-                examples_per_page=6
-            )
-        
-        submit_button = gr.Button("Detect Faces", variant="primary", scale=2)
-        
-        with gr.Column():
-            gr.Markdown("### Detection Controls")
-            model_choice = gr.Radio(["MMDetection"], value="MMDetection", visible=False)
-            task_type = gr.Dropdown(
-                ["Face Detection", "Face Pose Estimation", "Gaze Estimation [experimental]"],
-                value="Face Detection",
-                label="Select Task"
-            )
-            conf_threshold = gr.Slider(0.05, 0.95, 0.25, step=0.05, label="Confidence Threshold")
-            max_detections = gr.Slider(1, 10, 3, step=1, label="Max Detections")
+        # Examples
+        gr.Examples(
+            examples=[
+                ["images/allocebus_000003.jpeg"],
+                ["images/tarsius_000120.jpeg"],
+                ["images/nasalis_proboscis-monkey.png"],
+                ["images/macaca_000032.jpeg"],
+                ["images/mandrillus_000011.jpeg"],
+                ["images/pongo_000006.jpeg"]
+            ],
+            inputs=input_file
+        )
         
-        # Event handlers
-        def handle_file_change(file_obj):
-            """Handle file upload preview locally"""
-            if file_obj is None:
-                return gr.update(visible=False), gr.update(visible=False), gr.update(visible=True)
-            
-            try:
-                # Forward to backend for preview
-                result = client.predict(file_obj, fn_index=0)
-                return result[:3]  # Return first 3 outputs for preview
-            except:
-                # Fallback local preview
-                if file_obj.name.lower().endswith(('.png', '.jpg', '.jpeg', '.bmp')):
-                    return gr.update(value=file_obj.name, visible=True), gr.update(visible=False), gr.update(visible=False)
-                else:
-                    return gr.update(visible=False), gr.update(visible=False), gr.update(visible=True)
+        submit_button = gr.Button("Detect Faces", variant="primary")
         
-        def handle_webcam_change(img):
-            """Handle webcam capture"""
-            if img is None:
-                return gr.update(visible=False), gr.update(visible=True)
-            try:
-                # Save and forward to backend
-                with tempfile.NamedTemporaryFile(suffix='.png', delete=False) as tmp:
-                    img.save(tmp, 'PNG')
-                    result = client.predict(tmp.name, fn_index=1)
-                    os.unlink(tmp.name)
-                return result[:2]  # Return webcam-related outputs
-            except:
-                # Fallback
-                return gr.update(value=img, visible=True), gr.update(visible=False)
+        # Controls
+        model_choice = gr.Radio(["MMDetection"], value="MMDetection", visible=False)
+        task_type = gr.Dropdown(
+            ["Face Detection", "Face Pose Estimation", "Gaze Estimation [experimental]"],
+            value="Face Detection"
+        )
+        conf_threshold = gr.Slider(0.05, 0.95, 0.25, step=0.05, label="Confidence")
+        max_detections = gr.Slider(1, 10, 3, step=1, label="Max Detections")
         
-        # Wire events
+        # Simple preview handlers
         input_file.change(
-            handle_file_change,
+            lambda f: (gr.update(value=f, visible=bool(f)), gr.update(visible=False)) if f and f.name.endswith(('.jpg','.jpeg','.png')) else (gr.update(visible=False), gr.update(value=f, visible=bool(f))),
             inputs=[input_file],
-            outputs=[display_raw_image_file, display_raw_video_file, input_file]
+            outputs=[display_raw_image_file, display_raw_video_file]
         )
         
         input_webcam.change(
-            handle_webcam_change,
+            lambda img: gr.update(value=img, visible=bool(img)),
             inputs=[input_webcam],
-            outputs=[display_raw_image_webcam, input_webcam]
+            outputs=[display_raw_image_webcam]
         )
         
+        # Main processing
         submit_button.click(
             process_media,
             inputs=[input_file, input_webcam, model_choice, conf_threshold, max_detections, task_type],
@@ -176,8 +130,9 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
             ]
         )
         
-        clear_all_button.click(
-            clear_all,
+        # Clear all
+        clear_button.click(
+            lambda: [gr.update(value=None)] * 7,
             outputs=[
                 input_file,
                 input_webcam,