The official code of paper Adaptive Knowledge Distillation for High-Quality Unsupervised MRI Reconstruction with Model-Driven Priors.
train teacher models
python main.py --cfg-path opts/bit_parcel.yaml
train student model
python main.py --cfg-path opts/bit_dccnn2_weighted_distill_parcel.yaml
train teacher models with fastmri dataset
python main.py --cfg-path opts/scalefastmri_parcel.yaml
train student models with fastmri dataset
python main.py --cfg-path opts/scalefastmri_dccnn2_weighted_distill_parcel.yaml