Upload plant disease classifier model

README.md

@@ -0,0 +1,22 @@
+# Plant Disease Classification Model
+
+A PyTorch model for classifying plant diseases in Apple, Tomato, and Corn crops.
+
+## Model Details
+
+- **Model Type**: Image Classification
+- **Architecture**: EfficientNet-B2 with Attention
+- **Input Size**: 224x224 RGB images
+- **Output**: Disease classification
+
+## Usage
+
+```python
+from inference import load_model, predict_image
+
+# Load model
+model, class_names = load_model("model.pth")
+
+# Make prediction
+results = predict_image("your_image.jpg", model, class_names)
+print(results)

config.json

@@ -0,0 +1,25 @@
+{
+  "model_type": "image-classification",
+  "num_classes": 14,
+  "class_names": [
+    "Apple_Apple_Scab",
+    "Apple_Black_Rot",
+    "Apple_Cedar_Apple_Rust",
+    "Apple_Healthy",
+    "Corn_(maize)_Cercospora_Leaf_Spot",
+    "Corn_(maize)_Common_Rust_",
+    "Corn_(maize)_Healthy",
+    "Corn_(maize)_Northern_Leaf_Blight",
+    "Tomato_Bacterial_Spot",
+    "Tomato_Early_Blight",
+    "Tomato_Healthy",
+    "Tomato_Late_Blight",
+    "Tomato_Septoria_Leaf_Spot",
+    "Tomato_Yellow_Leaf_Curl_Virus"
+  ],
+  "input_size": [
+    224,
+    224
+  ],
+  "framework": "pytorch"
+}

inference.py

@@ -0,0 +1,87 @@
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torchvision import transforms
+from PIL import Image
+
+class PlantDiseaseClassifier(nn.Module):
+    def __init__(self, num_classes, dropout_rate=0.3):
+        super(PlantDiseaseClassifier, self).__init__()
+        
+        # Use EfficientNet as backbone
+        from torchvision import models
+        self.backbone = models.efficientnet_b2(pretrained=False)
+        
+        # Get feature dimension
+        num_features = self.backbone.classifier[1].in_features
+        
+        # Replace classifier with custom head
+        self.backbone.classifier = nn.Identity()
+        
+        # Attention mechanism
+        self.attention = nn.Sequential(
+            nn.AdaptiveAvgPool2d(1),
+            nn.Flatten(),
+            nn.Linear(num_features, num_features // 4),
+            nn.ReLU(),
+            nn.Linear(num_features // 4, num_features),
+            nn.Sigmoid()
+        )
+        
+        # Custom classifier head
+        self.classifier = nn.Sequential(
+            nn.Dropout(dropout_rate),
+            nn.Linear(num_features, 512),
+            nn.BatchNorm1d(512),
+            nn.ReLU(),
+            nn.Dropout(dropout_rate * 0.5),
+            nn.Linear(512, 256),
+            nn.BatchNorm1d(256),
+            nn.ReLU(),
+            nn.Dropout(dropout_rate * 0.3),
+            nn.Linear(256, num_classes)
+        )
+    
+    def forward(self, x):
+        features = self.backbone.features(x)
+        pooled = F.adaptive_avg_pool2d(features, 1)
+        pooled = torch.flatten(pooled, 1)
+        attention_weights = self.attention(features)
+        attended_features = pooled * attention_weights
+        output = self.classifier(attended_features)
+        return output
+
+def load_model(model_path):
+    checkpoint = torch.load(model_path, map_location='cpu')
+    num_classes = len(checkpoint['class_names'])
+    model = PlantDiseaseClassifier(num_classes=num_classes)
+    model.load_state_dict(checkpoint['model_state_dict'])
+    model.eval()
+    return model, checkpoint['class_names']
+
+def predict_image(image_path, model, class_names):
+    transform = transforms.Compose([
+        transforms.Resize((224, 224)),
+        transforms.ToTensor(),
+        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+    ])
+    
+    image = Image.open(image_path).convert('RGB')
+    image_tensor = transform(image).unsqueeze(0)
+    
+    with torch.no_grad():
+        outputs = model(image_tensor)
+        probabilities = F.softmax(outputs, dim=1)[0]
+    
+    # Get top predictions
+    top_probs, top_indices = torch.topk(probabilities, 3)
+    
+    results = []
+    for i in range(len(top_indices)):
+        results.append({
+            "label": class_names[top_indices[i].item()],
+            "score": top_probs[i].item()
+        })
+    
+    return results

model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a1511cc75d57f3e05f1d7ef48842e52e34edddf4eeb3537795068e3e1f2ebf8
+size 38649138