try to fix segm

app.py

@@ -49,7 +49,6 @@ COCO_DETECTION_CLASSES = [
 ]
 
 # COCO-Stuff (171 Classes)
-# FIX: Added 'unlabeled' at index 0 so 'person' aligns with Index 1
 COCO_STUFF_CLASSES = [
     "unlabeled", # Index 0 (Background)
     "person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", "traffic light",
@@ -101,7 +100,7 @@ def run_prediction(image, confidence_threshold, resolution, model_name):
     image_input = image.resize((resolution, resolution))
     
     if model_name in SEGMENTATION_MODELS:
-        return run_segmentation(model, image_input, image)
+        return run_segmentation(model, image_input)
     else:
         return run_detection(model, image_input, image, confidence_threshold)
 
@@ -149,7 +148,7 @@ def run_detection(model, image_input, original_image, confidence_threshold):
     
     return annotated, analytics_text, {"count": len(boxes), "objects": class_counts}
 
-
+# Segm code opt 1
 '''
 def run_segmentation(model, image_input, original_image):
     mask_tensor = model.predict(image_input)
@@ -198,17 +197,15 @@ def run_segmentation(model, image_input, original_image):
 
     return Image.fromarray(blended), analytics_text, {"classes_found": list(found_classes)}
 '''
+
+# Segm code opt 2
 def run_segmentation(model, image):
     """
     Handles Segmentation: Returns Tensor of shape (H, W) with class IDs.
     """
-    # 1. Run Prediction
     mask_tensor = model.predict(image)
-    
-    # 2. Convert to Numpy
     mask_np = mask_tensor.cpu().numpy().astype(np.uint8)
 
-    # 3. Create a Colored Mask
     h, w = mask_np.shape
     colored_mask = np.zeros((h, w, 3), dtype=np.uint8)
     
@@ -221,7 +218,6 @@ def run_segmentation(model, image):
         color = np.random.randint(50, 255, size=3) 
         colored_mask[mask_np == cls_id] = color
 
-    # 4. Blend with Original Image
     image_np = np.array(image)
     if image_np.shape[:2] != colored_mask.shape[:2]:
         colored_mask = cv2.resize(colored_mask, (image_np.shape[1], image_np.shape[0]), interpolation=cv2.INTER_NEAREST)
@@ -245,7 +241,6 @@ with gr.Blocks(theme=theme) as demo:
             
             with gr.Accordion("Settings", open=True):
                 conf_slider = gr.Slider(0.0, 1.0, value=0.4, step=0.05, label="Confidence (Detection Only)")
-                # BUMPED DEFAULT TO 640 for sharper masks
                 res_slider = gr.Slider(384, 1024, value=640, step=32, label="Inference Resolution")
                 
                 model_selector = gr.Dropdown(