metadata
{}
Channel Reconstruction U-Net
A U-Net model trained to reconstruct missing channels in multi-channel drone imagery.
Model Description
This model predicts missing spectral channels from available ones in 5-channel remote sensing images (224x224 patches).
Intended Use
- Reconstruct corrupted or missing drone image channels
- Fill gaps in multi-spectral data
- Evaluate spectral channel redundancy and relationships