Upload app.py

app.py

@@ -573,100 +573,33 @@ with gr.Blocks(theme=theme, title="Car Window Segmentation") as demo:
     Compare your custom trained YOLOv8 model against state-of-the-art Zero-Shot models!
     """)
 
-    # ── TAB 1: Custom Models ──
-    with gr.Tab("Test Custom Models"):
-        with gr.Row():
-            with gr.Column(scale=1):
-                input_image_custom = gr.Image(type="numpy", label="Upload Window Image")
-                model_dropdown_custom = gr.Dropdown(
-                    choices=[
-                        "YOLOv8x-seg (Custom Window)",
-                        "YOLOv8x-seg",
-                        "YOLO11x-seg",
-                        "SAM + YOLO (Strategy 1: Bbox + 5 Points)",
-                        "SAM + YOLO (Strategy 2: Mask + 5 Points)",
-                        "SAM + YOLO (Strategy 3: Direct Mask Prompting)",
-                        "Mask R-CNN",
-                        "SegFormer"
-                    ],
-                    value="SegFormer",
-                    label="Select Custom Model",
-                    info="Choose between fine-tuned models and experimental architectures"
-                )
-                morph_checkbox = gr.Checkbox(
-                    value=False,
-                    label="Apply Morphological Cleanup",
-                    info="Fills holes inside mask (Closing) and removes tiny noise blobs (Opening). Visual only — does not affect mAP metrics."
-                )
-                submit_btn_custom = gr.Button("Run Segmentation", variant="primary", size="lg")
-            with gr.Column(scale=1):
-                output_image_custom = gr.Image(label="Segmentation Result", interactive=False)
-                output_mask_custom = gr.Image(label="Binary Mask (White=Object, Black=Background)", interactive=False)
-                output_stats_custom = gr.Textbox(label="Detection Statistics", interactive=False)
 
-        if mirror_examples:
-            gr.Markdown("### Click any window image below to load it")
-            custom_gallery = gr.Gallery(value=mirror_examples, columns=10, height=120, object_fit="cover", allow_preview=False, show_label=False)
-            
-            def load_mirror_img_custom(evt: gr.SelectData):
-                return mirror_examples[evt.index]
-                
-            custom_gallery.select(fn=load_mirror_img_custom, inputs=None, outputs=input_image_custom)
 
-        submit_btn_custom.click(
-            fn=process_image,
-            inputs=[input_image_custom, model_dropdown_custom, gr.State(""), morph_checkbox],
-            outputs=[output_image_custom, output_mask_custom, output_stats_custom]
-        )
+    # ── TAB 3: Comprehensive Evaluation ──
+    with gr.Tab("Comprehensive Evaluation"):
+        gr.Markdown("### 🔍 Comprehensive Evaluation: Results from All Trained and Pretrained Models")
+        gr.Markdown("""**The following models will run and display their results below:**
 
-    # ── TAB 2: Pretrained Zero-Shot Models ──
-    with gr.Tab("Pretrained Zero-Shot Models"):
-        with gr.Row():
-            with gr.Column(scale=1):
-                input_image_pretrained = gr.Image(type="numpy", label="Upload Window Image")
-                model_dropdown_pretrained = gr.Dropdown(
-                    choices=[
-                        "Grounding DINO (Zero-Shot Detection)", 
-                        "Grounded SAM (Zero-Shot Segmentation)",
-                        "Intelliarts Car Parts (Detectron2)"
-                    ],
-                    value="Grounded SAM (Zero-Shot Segmentation)",
-                    label="Select Pretrained Model",
-                    info="Finds windows purely based on the text prompt you provide below!"
-                )
-                text_prompt = gr.Textbox(
-                    value="car window. car glass. windshield.",
-                    label="What to search for? (Text Prompt)",
-                    info="Be sure to separate terms with a period."
-                )
-                submit_btn_pretrained = gr.Button("Run Zero-Shot Detection", variant="primary", size="lg")
-            with gr.Column(scale=1):
-                output_image_pretrained = gr.Image(label="Segmentation Result", interactive=False)
-                output_mask_pretrained = gr.Image(label="Binary Mask", interactive=False)
-                output_stats_pretrained = gr.Textbox(label="Detection Statistics", interactive=False)
+**Custom Trained Models:**
 
-        if mirror_examples:
-            gr.Markdown("### Click any window image below to load it")
-            pretrained_gallery = gr.Gallery(value=mirror_examples, columns=10, height=120, object_fit="cover", allow_preview=False, show_label=False)
-            
-            def load_mirror_img_pretrained(evt: gr.SelectData):
-                return mirror_examples[evt.index]
-                
-            pretrained_gallery.select(fn=load_mirror_img_pretrained, inputs=None, outputs=input_image_pretrained)
+1. YOLOv8x-seg
+2. YOLO11x-seg
+3. Mask R-CNN
+4. SegFormer
+5. SegFormer + Morphological
+6. SAM + YOLO (Strategy 1, 2, 3)
 
-        submit_btn_pretrained.click(
-            fn=process_image,
-            inputs=[input_image_pretrained, model_dropdown_pretrained, text_prompt],
-            outputs=[output_image_pretrained, output_mask_pretrained, output_stats_pretrained]
-        )
+**Pretrained Zero-Shot Models:**
 
-    # ── TAB 3: Sequential Multi-Model Inference ──
-    with gr.Tab("Sequential Multi-Model Inference"):
-        gr.Markdown("### Upload an image to run through all our custom models sequentially!")
+1. Grounding DINO
+2. Grounded SAM
+3. Intelliarts Car Parts
+""")
+        
         with gr.Row():
             input_image_seq = gr.Image(type="numpy", label="Upload Window Image")
         with gr.Row():
-            submit_btn_seq = gr.Button("Run All Models in Sequence", variant="primary", size="lg")
+            submit_btn_seq = gr.Button("Run All Models", variant="primary", size="lg")
             
         if mirror_examples:
             gr.Markdown("### Or click any example image below to load it:")
@@ -675,6 +608,8 @@ with gr.Blocks(theme=theme, title="Car Window Segmentation") as demo:
             compare_gallery.select(fn=load_compare_img, inputs=None, outputs=input_image_seq)
 
         gr.Markdown("---")
+        gr.Markdown("## 🚀 Custom Trained Models")
+        
         gr.Markdown("### 1️⃣ YOLOv8x-seg")
         with gr.Row():
             seq_yolo_img = gr.Image(label="YOLO Overlay", interactive=False)
@@ -688,60 +623,114 @@ with gr.Blocks(theme=theme, title="Car Window Segmentation") as demo:
             seq_yolo11_bw = gr.Image(label="YOLO11x Binary Mask", interactive=False, image_mode="L")
         seq_yolo11_stats = gr.Textbox(label="YOLO11x Stats", interactive=False)
 
-            
         gr.Markdown("---")
-        gr.Markdown("### 2️⃣ Mask R-CNN (ResNet50-FPN)")
+        gr.Markdown("### 3️⃣ Mask R-CNN (ResNet50-FPN)")
         with gr.Row():
             seq_mrcnn_img = gr.Image(label="Mask R-CNN Overlay", interactive=False)
             seq_mrcnn_bw = gr.Image(label="Mask R-CNN Binary Mask", interactive=False, image_mode="L")
         seq_mrcnn_stats = gr.Textbox(label="Mask R-CNN Stats", interactive=False)
 
         gr.Markdown("---")
-        gr.Markdown("### 3️⃣ SegFormer (Transformer - Best Model)")
+        gr.Markdown("### 4️⃣ SegFormer (Transformer - Best Model)")
         with gr.Row():
             seq_segf_img = gr.Image(label="SegFormer Overlay", interactive=False)
             seq_segf_bw = gr.Image(label="SegFormer Binary Mask", interactive=False, image_mode="L")
         seq_segf_stats = gr.Textbox(label="SegFormer Stats", interactive=False)
 
         gr.Markdown("---")
-        gr.Markdown("### 4️⃣ SegFormer + Morphological Cleanup (Holes Filled + Sharp Borders)")
+        gr.Markdown("### 5️⃣ SegFormer + Morphological Cleanup (Holes Filled + Sharp Borders)")
         with gr.Row():
             seq_segf_morph_img = gr.Image(label="SegFormer + Morph Overlay", interactive=False)
             seq_segf_morph_bw = gr.Image(label="SegFormer + Morph Binary Mask", interactive=False, image_mode="L")
         seq_segf_morph_stats = gr.Textbox(label="SegFormer + Morph Stats", interactive=False)
 
-        def run_sequential(img):
-            if img is None: return [None]*15
+        gr.Markdown("---")
+        gr.Markdown("### 6️⃣ SAM + YOLO (Strategy 1: Bbox + 5 Points)")
+        with gr.Row():
+            seq_sam1_img = gr.Image(label="SAM+YOLO Strat 1 Overlay", interactive=False)
+            seq_sam1_bw = gr.Image(label="SAM+YOLO Strat 1 Binary Mask", interactive=False, image_mode="L")
+        seq_sam1_stats = gr.Textbox(label="SAM+YOLO Strat 1 Stats", interactive=False)
+
+        gr.Markdown("---")
+        gr.Markdown("### 7️⃣ SAM + YOLO (Strategy 2: Mask + 5 Points)")
+        with gr.Row():
+            seq_sam2_img = gr.Image(label="SAM+YOLO Strat 2 Overlay", interactive=False)
+            seq_sam2_bw = gr.Image(label="SAM+YOLO Strat 2 Binary Mask", interactive=False, image_mode="L")
+        seq_sam2_stats = gr.Textbox(label="SAM+YOLO Strat 2 Stats", interactive=False)
+
+        gr.Markdown("---")
+        gr.Markdown("### 8️⃣ SAM + YOLO (Strategy 3: Direct Mask Prompting)")
+        with gr.Row():
+            seq_sam3_img = gr.Image(label="SAM+YOLO Strat 3 Overlay", interactive=False)
+            seq_sam3_bw = gr.Image(label="SAM+YOLO Strat 3 Binary Mask", interactive=False, image_mode="L")
+        seq_sam3_stats = gr.Textbox(label="SAM+YOLO Strat 3 Stats", interactive=False)
+
+        gr.Markdown("---")
+        gr.Markdown("## 🌍 Pretrained Zero-Shot Models")
+
+        gr.Markdown("### 9️⃣ Grounding DINO (Zero-Shot Detection)")
+        with gr.Row():
+            seq_dino_img = gr.Image(label="Grounding DINO Overlay", interactive=False)
+            seq_dino_bw = gr.Image(label="Grounding DINO Binary Mask", interactive=False, image_mode="L")
+        seq_dino_stats = gr.Textbox(label="Grounding DINO Stats", interactive=False)
+
+        gr.Markdown("---")
+        gr.Markdown("### 🔟 Grounded SAM (Zero-Shot Segmentation)")
+        with gr.Row():
+            seq_gsam_img = gr.Image(label="Grounded SAM Overlay", interactive=False)
+            seq_gsam_bw = gr.Image(label="Grounded SAM Binary Mask", interactive=False, image_mode="L")
+        seq_gsam_stats = gr.Textbox(label="Grounded SAM Stats", interactive=False)
+
+        gr.Markdown("---")
+        gr.Markdown("### 1️⃣1️⃣ Intelliarts Car Parts (Detectron2)")
+        with gr.Row():
+            seq_intell_img = gr.Image(label="Intelliarts Car Parts Overlay", interactive=False)
+            seq_intell_bw = gr.Image(label="Intelliarts Car Parts Binary Mask", interactive=False, image_mode="L")
+        seq_intell_stats = gr.Textbox(label="Intelliarts Car Parts Stats", interactive=False)
+
+        def run_all_models(img):
+            if img is None: return [None]*33
             
-            # Run YOLO
             yolo_out, yolo_mask, yolo_stats = process_image(img, "YOLOv8x-seg", "", False)
-            
-            # Run YOLO11x
             yolo11_out, yolo11_mask, yolo11_stats = process_image(img, "YOLO11x-seg", "", False)
-            
-            # Run Mask R-CNN
             mrcnn_out, mrcnn_mask, mrcnn_stats = process_image(img, "Mask R-CNN", "", False)
-            
-            # Run SegFormer (Plain)
             segf_out, segf_mask, segf_stats = run_segformer(img, morph_cleanup=False)
-            
-            # Run SegFormer + Morphological Cleanup (Holes filled + Sharp borders)
             segf_morph_out, segf_morph_mask, segf_morph_stats = run_segformer(img, morph_cleanup=True)
             
+            sam1_out, sam1_mask, sam1_stats = process_image(img, "SAM + YOLO (Strategy 1: Bbox + 5 Points)", "", False)
+            sam2_out, sam2_mask, sam2_stats = process_image(img, "SAM + YOLO (Strategy 2: Mask + 5 Points)", "", False)
+            sam3_out, sam3_mask, sam3_stats = process_image(img, "SAM + YOLO (Strategy 3: Direct Mask Prompting)", "", False)
+            
+            dino_out, dino_mask, dino_stats = process_image(img, "Grounding DINO (Zero-Shot Detection)", "car window. car glass. windshield.", False)
+            gsam_out, gsam_mask, gsam_stats = process_image(img, "Grounded SAM (Zero-Shot Segmentation)", "car window. car glass. windshield.", False)
+            intell_out, intell_mask, intell_stats = process_image(img, "Intelliarts Car Parts (Detectron2)", "", False)
+            
             return (yolo_out, yolo_mask, yolo_stats,
                     yolo11_out, yolo11_mask, yolo11_stats,
                     mrcnn_out, mrcnn_mask, mrcnn_stats, 
                     segf_out, segf_mask, segf_stats,
-                    segf_morph_out, segf_morph_mask, segf_morph_stats)
+                    segf_morph_out, segf_morph_mask, segf_morph_stats,
+                    sam1_out, sam1_mask, sam1_stats,
+                    sam2_out, sam2_mask, sam2_stats,
+                    sam3_out, sam3_mask, sam3_stats,
+                    dino_out, dino_mask, dino_stats,
+                    gsam_out, gsam_mask, gsam_stats,
+                    intell_out, intell_mask, intell_stats)
 
         submit_btn_seq.click(
-            fn=run_sequential,
+            fn=run_all_models,
             inputs=[input_image_seq],
             outputs=[seq_yolo_img, seq_yolo_bw, seq_yolo_stats, 
                      seq_yolo11_img, seq_yolo11_bw, seq_yolo11_stats,
                      seq_mrcnn_img, seq_mrcnn_bw, seq_mrcnn_stats, 
                      seq_segf_img, seq_segf_bw, seq_segf_stats,
-                     seq_segf_morph_img, seq_segf_morph_bw, seq_segf_morph_stats]
+                     seq_segf_morph_img, seq_segf_morph_bw, seq_segf_morph_stats,
+                     seq_sam1_img, seq_sam1_bw, seq_sam1_stats,
+                     seq_sam2_img, seq_sam2_bw, seq_sam2_stats,
+                     seq_sam3_img, seq_sam3_bw, seq_sam3_stats,
+                     seq_dino_img, seq_dino_bw, seq_dino_stats,
+                     seq_gsam_img, seq_gsam_bw, seq_gsam_stats,
+                     seq_intell_img, seq_intell_bw, seq_intell_stats]
         )
 
 if __name__ == "__main__":