Update app.py

app.py

@@ -208,49 +208,72 @@ def validate_video(path):
         return False, str(e)
 
 # Input Resolution
-def resolve_input(mode, uploaded_img, uploaded_imgs, uploaded_vid, url):
+def resolve_input(mode, uploaded_files, url):
     """
-    Resolves the input source based on user selection.
-    Supports single image, multiple images, video, or URL-based media.
-
+    Resolves the media input based on selected mode.
+    - If mode is 'Upload', accepts either:
+        * 1–5 images (PIL.Image)
+        * OR 1 video file (file path as string)
+    - If mode is 'URL', fetches remote image or video.
+    
     Args:
-        mode (str): Input mode - 'Upload' or 'URL'.
-        uploaded_img (PIL.Image.Image): Single uploaded image.
-        uploaded_imgs (List[PIL.Image.Image]): List of uploaded images (batch).
-        uploaded_vid (str): Uploaded video file path.
-        url (str): URL pointing to media content.
+        mode (str): 'Upload' or 'URL'
+        uploaded_files (List[Union[PIL.Image.Image, str]]): Uploaded media
+        url (str): URL to image or video
 
     Returns:
-        List[Union[PIL.Image.Image, str, None]]: A list of media items to process.
+        List[Union[PIL.Image.Image, str]] or None
     """
-    logger.info(f"Resolving input based on mode: {mode}")
     try:
+        logger.info(f"Resolving input for mode: {mode}")
+
         if mode == "Upload":
-            # Prefer batch if provided
-            if uploaded_imgs and len(uploaded_imgs) > 0:
-                return uploaded_imgs
-            elif uploaded_img:
-                return [uploaded_img]
-            elif uploaded_vid:
-                return [uploaded_vid]
-            else:
-                logger.warning("No valid upload provided.")
+            if not uploaded_files:
+                logger.warning("No upload detected.")
+                return None
+
+            image_files = [f for f in uploaded_files if isinstance(f, Image.Image)]
+            video_files = [f for f in uploaded_files if isinstance(f, str) and f.lower().endswith((".mp4", ".mov", ".avi"))]
+
+            if image_files and video_files:
+                logger.warning("Mixed media upload not supported (images + video).")
+                return None
+
+            if image_files:
+                if 1 <= len(image_files) <= 5:
+                    logger.info(f"Accepted {len(image_files)} image(s).")
+                    return image_files
+                logger.warning("Invalid number of images. Must be 1 to 5.")
+                return None
+
+            if video_files:
+                if len(video_files) == 1:
+                    logger.info("Accepted single video upload.")
+                    return video_files
+                logger.warning("Only one video allowed.")
                 return None
 
+            logger.warning("Unsupported upload type.")
+            return None
+
         elif mode == "URL":
-            media_from_url = fetch_media_from_url(url)
-            if media_from_url:
-                return [media_from_url]
+            if not url:
+                logger.warning("URL mode selected but URL is empty.")
+                return None
+            media = fetch_media_from_url(url)
+            if media:
+                logger.info("Media successfully fetched from URL.")
+                return [media]
             else:
-                logger.warning("Failed to fetch valid media from URL.")
+                logger.warning("Failed to resolve media from URL.")
                 return None
 
         else:
-            logger.warning("Invalid input mode selected.")
+            logger.error(f"Invalid mode selected: {mode}")
             return None
 
     except Exception as e:
-        logger.error(f"Error resolving input: {e}")
+        logger.error(f"Exception in resolve_input(): {e}")
         return None
 
 @timeout_decorator.timeout(35, use_signals=False)  # 35 sec limit per image
@@ -372,7 +395,7 @@ def process_image(
         return None, {"error": str(e)}, None
 
 # Main Handler
-def handle(mode, img, imgs, vid, url, run_det, det_model, det_confidence, run_seg, seg_model, run_depth, depth_model, blend):
+def handle(mode, uploaded_files, url, run_det, det_model, det_confidence, run_seg, seg_model, run_depth, depth_model, blend):
     """
     Master handler for resolving input and processing.
     Returns outputs for Gradio interface.
@@ -381,7 +404,7 @@ def handle(mode, img, imgs, vid, url, run_det, det_model, det_confidence, run_se
     logger.info(f"Session ID: {session_id} | Handler activated with mode: {mode}")
     start_time = time.time()
 
-    media = resolve_input(mode, img, imgs, vid, url)
+    media = resolve_input(mode, uploaded_files, url)
     if not media:
         return None, format_error("No valid input provided. Please check your upload or URL."), None
 
@@ -416,12 +439,28 @@ def handle(mode, img, imgs, vid, url, run_det, det_model, det_confidence, run_se
 with gr.Blocks() as demo:
     gr.Markdown("## Unified Visual Intelligence System (UVIS)")
 
-    # Input Mode Selection
+    # Input Mode Toggle
     mode = gr.Radio(["Upload", "URL"], value="Upload", label="Input Mode")
-    img = gr.Image(type="pil", label="Upload Image")
-    imgs = gr.Gallery(label="Upload Multiple Images (Up to 5)")
-    vid = gr.Video(label="Upload Video (<= 30s)")
-    url = gr.Textbox(label="URL (Image/Video)")
+    
+    # File upload: accepts multiple images or one video (user chooses wisely)
+    media_upload = gr.File(
+        label="Upload Images (1–5) or 1 Video",
+        file_types=["image", ".mp4", ".mov", ".avi"],
+        file_count="multiple"
+    )
+    
+    # URL input
+    url = gr.Textbox(label="URL (Image/Video)", visible=False)
+    
+    # Toggle visibility
+    def toggle_inputs(selected_mode):
+        return [
+            gr.update(visible=(selected_mode == "Upload")),  # media_upload
+            gr.update(visible=(selected_mode == "URL"))      # url
+        ]
+    
+    mode.change(toggle_inputs, inputs=mode, outputs=[media_upload, url])
+
 
     # Task Selection with parameters
     with gr.Accordion("Object Detection Settings", open=False):
@@ -458,7 +497,7 @@ with gr.Blocks() as demo:
     # Button Click Event
     run.click(
         handle,
-        inputs=[mode, img, imgs, vid, url, run_det, det_model, det_confidence, run_seg, seg_model, run_depth, depth_model, blend],
+        inputs=[mode, media_upload, url, run_det, det_model, det_confidence, run_seg, seg_model, run_depth, depth_model, blend],
         outputs=[img_out, json_out, zip_out]
     )