research
∙
05/26/2023
Improving Knowledge Distillation via Regularizing Feature Norm and Direction
Knowledge distillation (KD) exploits a large well-trained model (i.e., t...
share
research
∙
04/06/2022
