You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Deep learning-based semantic segmentation of Sentinel-2 multispectral imagery for agricultural land cover classification. The project uses a custom Attention U-Net with spectral indices (NDVI, NDWI, MNDWI, EVI, AWEI) and an end-to-end pipeline including auto-labeling, patch-based training, and GIS-ready inference.