Stochastic Ghost Batch for Self-distillation with Dynamic Soft Label
Authors:
- Qian LiQLQian Li
- Qingyuan HuQHQingyuan Hu
- Saiyu QiSQSaiyu Qi
- Yong QiYQYong Qi
- Di WuDWDi Wu
- Yun LinYLYun Lin
- Jin Song DongJDJin Song Dong
- 19 January 2022
- journal article
- research article
- Published by Elsevier in Knowledge-Based Systems
- Vol. 241
Abstract
No abstract availableKeywords
Funding Information
- National Key Research and Development Program of China
- National Key Research and Development Program of China Stem Cell and Translational Research (2020AAA0107702)
- National Natural Science Foundation of China (61672421)
- Ministry of Education of the People's Republic of China (2020 KJ010801)
This publication has 12 references indexed in Scilit:
- MixUp as Locally Linear Out-of-Manifold RegularizationProceedings of the AAAI Conference on Artificial Intelligence, 2019
- Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness Against Adversarial Attack2019
- Deep Metric Learning Beyond Binary Supervision2019
- Towards Evaluating the Robustness of Neural Networks2017
- Identity Mappings in Deep Residual Networks2016
- Cross Modal Distillation for Supervision Transfer2016
- Deep Residual Learning for Image Recognition2016
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- Lipschitz Behavior of the Robust RegularizationSIAM Journal on Control and Optimization, 2010
- Model compression2006