dr-disease-mobilenet-v2

🩺 Dr. Disease (MobileNetV2)

Dr. Disease is a lightweight, offline-capable image classification model designed to detect crop diseases on mobile devices. It is the core AI engine of the AgroTech Ecosystem.

🌾 Model Details

• Architecture: MobileNetV2 (Transfer Learning)
• Framework: TensorFlow / Keras
• Input Resolution: 224x224 pixels
• Target Platform: Android/iOS (via TensorFlow Lite)
• License: Apache 2.0

🎯 Intended Use

• Primary Use Case: Real-time diagnosis of crop diseases (e.g., Tomato Early Blight, Potato Late Blight) from leaf photos.
• Users: Farmers and agricultural extension workers in low-bandwidth regions.

📊 Performance

• Dataset: Trained on the PlantVillage Dataset.
• Goal Accuracy: >90% on validation set.
• Latency: Optimized for <200ms inference time on standard smartphones.

⚠️ Limitations

• Lighting conditions significantly affect accuracy.
• Currently limited to 10 common crop diseases (list to be updated).

---

Part of the AgroTech Project.