An example of a UNet architecture built using torch for semantic segmentation.
The overall codebase is not well-refined but to showcase using torch framework to built a U-Net
The model was trained for 5 epochs:
Example inference on test set:
- Add remaining inference script
- Add custom assets such as dataset used for training and inference