Update app.py

app.py

@@ -9,9 +9,10 @@ from pytorch_grad_cam import GradCAM, HiResCAM, ScoreCAM, GradCAMPlusPlus, Ablat
 from pytorch_grad_cam.utils.model_targets import ClassifierOutputTarget
 from pytorch_grad_cam.utils.image import show_cam_on_image
 from timm.data import create_transform
+from timm.data import infer_imagenet_subset, ImageNetInfo
 
 # List of available timm models
-MODELS = timm.list_models()
+MODELS = timm.list_pretrained()
 
 # List of available GradCAM methods
 CAM_METHODS = {
@@ -25,6 +26,16 @@ CAM_METHODS = {
     "FullGrad": FullGrad
 }
 
+class CustomDatasetInfo:
+    def __init__(self, label_names, label_descriptions=None):
+        self.label_names = label_names
+        self.label_descriptions = label_descriptions or label_names
+
+    def index_to_description(self, index, detailed=False):
+        if detailed and self.label_descriptions:
+            return self.label_descriptions[index]
+        return self.label_names[index]
+
 def load_model(model_name):
     model = timm.create_model(model_name, pretrained=True)
     model.eval()
@@ -50,9 +61,14 @@ def process_image(image_path, model):
     tensor = transform(image).unsqueeze(0)
     return tensor
 
-def get_cam_image(model, image, target_layer, cam_method):
+def get_cam_image(model, image, target_layer, cam_method, target_class):
+    if target_class is not None and target_class != "highest scoring":
+        target = ClassifierOutputTarget(target_class)
+    else:
+        target = None
+    
     cam = CAM_METHODS[cam_method](model=model, target_layers=[target_layer])
-    grayscale_cam = cam(input_tensor=image)
+    grayscale_cam = cam(input_tensor=image, targets=[target] if target else None)
     
     config = model.pretrained_cfg
     mean = torch.tensor(config['mean']).view(3, 1, 1)
@@ -79,20 +95,35 @@ def get_target_layer(model, target_layer_name):
         print(f"WARNING: Layer '{target_layer_name}' not found in the model.")
         return None
 
-def explain_image(model_name, image_path, cam_method, feature_module):
+def get_class_names(model):
+    dataset_info = None
+    label_names = model.pretrained_cfg.get("label_names", None)
+    label_descriptions = model.pretrained_cfg.get("label_descriptions", None)
+    if label_names is None:
+        imagenet_subset = infer_imagenet_subset(model)
+        if imagenet_subset:
+            dataset_info = ImageNetInfo(imagenet_subset)
+        else:
+            label_names = [f"LABEL_{i}" for i in range(model.num_classes)]
+    if dataset_info is None:
+        dataset_info = CustomDatasetInfo(
+            label_names=label_names,
+            label_descriptions=label_descriptions,
+        )
+    return dataset_info
+
+def explain_image(model_name, image_path, cam_method, feature_module, target_class):
     model = load_model(model_name)
     image = process_image(image_path, model)
     
     target_layer = get_target_layer(model, feature_module)
     
     if target_layer is None:
-        # Fallback to the last feature module or last convolutional layer
         feature_info = get_feature_info(model)
         if feature_info:
             target_layer = get_target_layer(model, feature_info[-1])
             print(f"Using last feature module: {feature_info[-1]}")
         else:
-            # Fallback to finding last convolutional layer
             for name, module in reversed(list(model.named_modules())):
                 if isinstance(module, torch.nn.Conv2d):
                     target_layer = module
@@ -102,17 +133,35 @@ def explain_image(model_name, image_path, cam_method, feature_module):
     if target_layer is None:
         raise ValueError("Could not find a suitable target layer.")
     
-    cam_image = get_cam_image(model, image, target_layer, cam_method)
-    return cam_image
+    target_class_index = None if target_class == "highest scoring" else int(target_class.split(':')[0])
+    cam_image = get_cam_image(model, image, target_layer, cam_method, target_class_index)
+    
+    with torch.no_grad():
+        out = model(image)
+    probabilities = out.squeeze(0).softmax(dim=0)
+    values, indices = torch.topk(probabilities, 5)  # Top 5 predictions
+    dataset_info = get_class_names(model)
+    labels = [
+        f"{i}: {dataset_info.index_to_description(i.item(), detailed=True)} ({v.item():.2%})"
+        for i, v in zip(indices, values)
+    ]
+    
+    return cam_image, "\n".join(labels)
 
 def update_feature_modules(model_name):
     model = load_model(model_name)
     feature_modules = get_feature_info(model)
     return gr.Dropdown(choices=feature_modules, value=feature_modules[-1] if feature_modules else None)
 
+def update_class_dropdown(model_name):
+    model = load_model(model_name)
+    dataset_info = get_class_names(model)
+    class_names = ["highest scoring"] + [f"{i}: {dataset_info.index_to_description(i, detailed=True)}" for i in range(model.num_classes)]
+    return gr.Dropdown(choices=class_names, value="highest scoring")
+
 with gr.Blocks() as demo:
-    gr.Markdown("# Explainable AI with timm models")
-    gr.Markdown("Upload an image, select a model, CAM method, and optionally a specific feature module to visualize the explanation.")
+    gr.Markdown("# Explainable AI with timm models. NOTE: This is a WIP but some models are functioning.")
+    gr.Markdown("Upload an image, select a model, CAM method, and optionally a specific feature module and target class to visualize the explanation.")
     
     with gr.Row():
         with gr.Column():
@@ -120,17 +169,20 @@ with gr.Blocks() as demo:
             image_input = gr.Image(type="filepath", label="Upload Image")
             cam_method_dropdown = gr.Dropdown(choices=list(CAM_METHODS.keys()), label="Select CAM Method")
             feature_module_dropdown = gr.Dropdown(label="Select Feature Module (optional)")
+            class_dropdown = gr.Dropdown(label="Select Target Class (optional)")
             explain_button = gr.Button("Explain Image")
         
         with gr.Column():
             output_image = gr.Image(type="pil", label="Explained Image")
+            prediction_text = gr.Textbox(label="Top 5 Predictions")
     
     model_dropdown.change(fn=update_feature_modules, inputs=[model_dropdown], outputs=[feature_module_dropdown])
+    model_dropdown.change(fn=update_class_dropdown, inputs=[model_dropdown], outputs=[class_dropdown])
     
     explain_button.click(
         fn=explain_image,
-        inputs=[model_dropdown, image_input, cam_method_dropdown, feature_module_dropdown],
-        outputs=[output_image]
+        inputs=[model_dropdown, image_input, cam_method_dropdown, feature_module_dropdown, class_dropdown],
+        outputs=[output_image, prediction_text]
     )
 
 demo.launch()