Update svm_vgg_preprocessor.py

svm_vgg_preprocessor.py

@@ -7,17 +7,14 @@ from PIL import Image
 
 class FeatureExtractor:
     def __init__(self):
-        # Load pretrained VGG16
+        # Load VGG16 with original architecture
         self.vgg = vgg16(weights='DEFAULT')
         self.vgg.eval()
         
-        # Feature extractors
-        self.conv_features = self.vgg.features
-        self.fc_features = torch.nn.Sequential(
-            *list(self.vgg.classifier.children())[:-2]
-        )
+        # Use only convolutional features (matches original code)
+        self.conv_extractor = self.vgg.features
         
-        # Preprocessing
+        # Original preprocessing
         self.preprocess = transforms.Compose([
             transforms.Resize((224, 224)),
             transforms.ToTensor(),
@@ -28,33 +25,27 @@ class FeatureExtractor:
         ])
 
     def extract_fc_cnn_features(self, image_path):
-        """Correct FC-CNN feature extraction"""
-        # Load and preprocess image
+        """Matches original FC-CNN feature extraction (conv features)"""
         img = Image.open(image_path).convert('RGB')
         img_tensor = self.preprocess(img).unsqueeze(0)
         
         with torch.no_grad():
-            # Get convolutional features
-            conv_out = self.conv_features(img_tensor)
-            # Flatten for FC layers
-            flattened = torch.flatten(conv_out, 1)
-            # Get FC features
-            fc_features = self.fc_features(flattened)
+            features = self.conv_extractor(img_tensor)
             
-        return fc_features.squeeze().numpy().flatten()
+        return features.squeeze().numpy().flatten()
 
     def extract_fv_cnn_features(self, image_path):
-        """FV-CNN feature extraction (unchanged)"""
+        """Matches original FV-CNN feature extraction (conv features)"""
         img = Image.open(image_path).convert('RGB')
         img_tensor = self.preprocess(img).unsqueeze(0)
         
         with torch.no_grad():
-            conv_features = self.conv_features(img_tensor)
+            features = self.conv_extractor(img_tensor)
             
-        return conv_features.squeeze().numpy().flatten()
+        return features.squeeze().numpy().flatten()
 
     def extract_combined_features(self, image_path):
-        """Combine both feature types"""
+        """Original concatenation of identical features"""
         fc = self.extract_fc_cnn_features(image_path)
         fv = self.extract_fv_cnn_features(image_path)
         return np.concatenate((fc, fv))