Python package for CNN pipelines
Four folders contains different steps for the workflow:
1 ./data_generate contains scripts for geneating 2-D near field, far field and traditional Mie scattering simulations.
2 ./data_loader constains scripts for loading, splitting, cropping and adding noise to the raw data.
3 ./model_training contains scripts for training different types of CNN, multi-training pipeline and noise-sensitivity test pipeline.
4 ./model_evaluation contains scripts that plot 3-D visualization of the accuracy for the trained model, or geneates an animation for the dataset.
Please pull chis repository and include it in Python path before running any script.