Multi-headed CNN for simultaneous potato/tomato classification (99.9% accuracy) and quality assessment (98.5% accuracy). Features Grad-CAM explainability, Streamlit interface, and 30% parameter reduction. Built for precision agriculture with real-time crop monitoring.
python opencv deep-learning tensorflow cnn quality-assessment agriculture-research multi-task-learning crop-classification disease-detection streamlit-webapp potato-disease tomato-disease
-
Updated
Jan 14, 2026 - Jupyter Notebook