HIGHLIGHTS
- who: JU YEON KANG and collaborators from the of Electrical Computer Engineering, Sungkyunkwan University, have published the paper: Binarized Neural Network With Parameterized Weight Clipping and Quantization Gap Minimization for Online Knowledge Distillation, in the Journal: (JOURNAL)
- what: In this study two performance improvements are proposed for online KD when a BNN is applied as a network: parameterized weight clipping (PWC) to reduce dead weights in the network quantization gap-aware adaptive temperature scheduling between the teacher networks. Next, the authors propose a method to alleviate the capacity shortage of binarized student networks caused by . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.