- dataset.py: Create training, validation and test samples in pytorch format.
- models.py:Create models, including ResNet101, ResNet152, Wide ResNet50, Wide ResNet101.
- train.py: Training of GSCNN.
- train_models.py: Train of GSCNNs based on four ResNets.
- evaluation.py: Evaluation of model performance.
- plot.py: Plotting of learning curve.
- predict.py: Prediction for BSV images in China.
- utils.py: Some utilities, including configuration and data augmentation pipelines.