No Data Augmentation? Alternative Regularizations for Effective Training on Small Datasets

Open in new window