V0

app.py

@@ -133,11 +133,24 @@ def setup_damage_detector(model_path, threshold=0.7):
         import traceback
         traceback.print_exc()
         return None, None
+
 def load_vit_deepfake_model(model_path, device):
     """Load the AutoAdapter model for deepfake detection"""
     try:
-        # Import AutoAdapterModel
-        from adapters import AutoAdapterModel
+        # First try to install the adapter-transformers library if not present
+        try:
+            import importlib.util
+            if importlib.util.find_spec("transformers") is None:
+                print("Installing transformers...")
+                os.system("pip install transformers")
+            if importlib.util.find_spec("adapter_transformers") is None:
+                print("Installing adapter-transformers...")
+                os.system("pip install -U adapter-transformers")
+        except Exception as e:
+            print(f"Warning: Could not install adapter-transformers: {e}")
+            
+        # Import AutoAdapterModel from adapter_transformers
+        from adapter_transformers import AutoAdapterModel
         
         # Create model from pretrained base
         print(f"Loading AutoAdapter model for deepfake detection...")
@@ -153,11 +166,36 @@ def load_vit_deepfake_model(model_path, device):
     except Exception as e:
         print(f"Error loading AutoAdapter model: {e}")
         traceback.print_exc()
-        return None
+        return None    
+   
 
 
 def preprocess_for_vit(image, device):
-    """Preprocess an image for AutoAdapter model deepfake detection"""
+    """Preprocess an image for adapter-transformers model deepfake detection"""
+    try:
+        # Import AutoProcessor from adapter_transformers 
+        from adapter_transformers import AutoTokenizer, AutoFeatureExtractor
+        
+        # Get the processor for the adapter model - using feature extractor for vision models
+        processor = AutoFeatureExtractor.from_pretrained("Askhedi/Car_fraud_damage_detector")
+        
+        # Ensure image is RGB
+        if len(image.shape) == 3 and image.shape[2] == 3:
+            if image.dtype != np.uint8:
+                image = (image * 255).astype(np.uint8)
+            rgb_img = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
+        else:
+            rgb_img = image
+            
+        # Use the processor to transform the image
+        inputs = processor(images=Image.fromarray(rgb_img), return_tensors="pt")
+        inputs = {k: v.to(device) for k, v in inputs.items()}
+        
+        return inputs
+    except Exception as e:
+        print(f"Error preprocessing image for adapter model: {e}")
+        traceback.print_exc()
+        return None    """Preprocess an image for AutoAdapter model deepfake detection"""
     try:
         # Import AutoProcessor
         from transformers import AutoProcessor
@@ -244,7 +282,7 @@ def detect_damage(img, damage_detector):
         return None, None, []
 
 def check_deepfake_vit(image, damage_regions, deepfake_model, device, threshold=0.5):
-    """Check if damage regions are deepfakes using AutoAdapter model"""
+    """Check if damage regions are deepfakes using adapter-transformers model"""
     results = []
     
     if deepfake_model is None:
@@ -378,22 +416,39 @@ def visualize_results(image, damage_outputs, deepfake_results, damage_threshold)
 def process_image(input_image, damage_model_path, deepfake_model_path, damage_threshold, 
                  deepfake_threshold, skip_damage, device_str, usage_count):
     """Process an image through the detection pipeline"""
+    # Handle empty or None usage_count
+    if usage_count is None:
+        usage_count = 0
+    
+    # Ensure usage_count is an integer
+    try:
+        usage_count = int(usage_count)
+    except (TypeError, ValueError):
+        usage_count = 0
+        
     # Increment usage count and check if limit reached
     usage_count = usage_count + 1
     
     progress_info = []
     progress_info.append(f"Usage: {usage_count}/{MAX_TRIES}")
     
+    # Check if usage limit reached
+    if usage_count > MAX_TRIES:
+        return None, f"⚠️ You have reached the maximum number of tries allowed ({MAX_TRIES}).", usage_count
+    
     # Check model files
     if not skip_damage and damage_model_path:
         if not os.path.exists(damage_model_path):
             progress_info.append(f"ERROR: Damage model not found at {damage_model_path}")
     
     if deepfake_model_path and not os.path.exists(deepfake_model_path):
-        progress_info.append(f"ERROR: ViT deepfake model not found at {deepfake_model_path}")
+        progress_info.append(f"ERROR: Deepfake model not found at {deepfake_model_path}")
     
     # Convert image to proper format
     try:
+        if input_image is None:
+            return None, "Please upload an image to analyze.", usage_count
+            
         if isinstance(input_image, dict) and "path" in input_image:
             img = cv2.imread(input_image["path"])
         elif isinstance(input_image, str):
@@ -427,14 +482,14 @@ def process_image(input_image, damage_model_path, deepfake_model_path, damage_th
         else:
             progress_info.append("❌ Failed to initialize damage detector")
         
-    # Setup ViT deepfake detector
+    # Setup deepfake detector
     if deepfake_model_path:
-        progress_info.append("Setting up ViT deepfake detector...")
+        progress_info.append("Setting up deepfake detector...")
         deepfake_model = load_vit_deepfake_model(deepfake_model_path, device)
         if deepfake_model:
-            progress_info.append("✅ ViT deepfake detector initialized")
+            progress_info.append("✅ Deepfake detector initialized")
         else:
-            progress_info.append("❌ Failed to initialize ViT deepfake detector")
+            progress_info.append("❌ Failed to initialize deepfake detector")
      
     # Step 1: Detect damage
     progress_info.append("Detecting damaged regions...")
@@ -447,10 +502,10 @@ def process_image(input_image, damage_model_path, deepfake_model_path, damage_th
     else:
         progress_info.append("No damage regions detected")
     
-    # Step 2: Check for deepfakes using ViT
+    # Step 2: Check for deepfakes
     deepfake_results = []
     if deepfake_model is not None:
-        progress_info.append("Analyzing regions for deepfakes using ViT model...")
+        progress_info.append("Analyzing regions for deepfakes...")
         start_time = time.time()
         deepfake_results = check_deepfake_vit(
             img, damage_regions, deepfake_model, device, float(deepfake_threshold)
@@ -505,7 +560,7 @@ def process_image(input_image, damage_model_path, deepfake_model_path, damage_th
         progress_info.append(f"\nRemaining tries: {MAX_TRIES - usage_count}")
     
     return result_img, "\n".join(progress_info), usage_count
-
+    
 def auto_install_dependencies():
     """Attempt to install dependencies if needed"""
     try:
@@ -560,7 +615,6 @@ def create_gradio_interface():
     )
     
     # Initialize usage counter
-    usage_counter = gr.State(0)
     
     with gr.Blocks(title="Car Damage & Deepfake Detector", theme=theme) as app:
         gr.Markdown("""
@@ -574,7 +628,8 @@ def create_gradio_interface():
         
         ⚠️ **Note: You have a maximum of 3 tries to analyze images.**
         """)
-        
+        usage_counter = gr.State(0)
+
         # Main Interface Tab
         with gr.Tab("Analyze Image"):
             with gr.Row():

requirements.txt

@@ -7,4 +7,6 @@ gradio>=3.50.0
 python-box>=7.1.1
 fvcore>=0.1.5.post20221221; platform_system!="Darwin"
 iopath>=0.1.9; platform_system!="Darwin"
-pycocotools>=2.0.6; platform_system!="Darwin"
+pycocotools>=2.0.6; platform_system!="Darwin"
+adapters
+transformers