leaf-disease

🌿 Leaf Disease Classification with EfficientNetB0

Leaf-Disease-Classification-EfficientNetB0 is a deep learning-powered app that identifies 38 types of plant diseases from leaf images using a fine-tuned EfficientNetB0 model trained on the New Plant Diseases Dataset.

📸 Upload an image or select an example below. Get instant predictions on plant disease type — with confidence score and class label.

---

🌐 Demo

Deployed on Hugging Face Spaces:

---

🧠 Model Details

• Model: EfficientNetB0 (Transfer Learning from ImageNet)
• Classes: 38 plant disease types (Apple, Corn, Potato, Tomato, etc.)
• Input Size: 224×224
• Preprocessing: EfficientNet preprocess_input
• Training:
  • Optimizer: Adam
  • Loss Function: Categorical Crossentropy
  • Epochs: 20
• Validation Accuracy: ~97%

---

📓 Training Notebook (Kaggle)

The model was trained using TensorFlow and Keras on Kaggle.
🔗 leaf-disease-classification-efficientnetb0-new1-1
Includes:

• Data Augmentation (Shear, Zoom, Flip)
• Early Stopping & Model Checkpointing
• Confusion Matrix Visualization
• Final .keras weights file exported

---

Features

• Classifies 38 different plant diseases
• Supports image upload and drag-and-drop
• Outputs top 3 predicted classes with confidence scores
• Lightweight & fast inference — powered by Gradio

---

📁 Folder Structure

Leaf-Disease-Project-CE180166/
├── app.py                   # Gradio interface main script
├── inference_utils.py       # Helper functions (Load model & Predict)
├── model/
│   └── final_model_20epochs.keras  # Trained Model Weights
├── assets/                  # Demo/test images
│   ├── Potato__Early_blight.jpg
│   └── ...
├── requirements.txt         # Dependencies
└── README.md                # Project Documentation