AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

MixSKD: Self-Knowledge Distillation from Mixup for Image Recognition

Chuanguang Yang, Zhulin An, Helong Zhou, Linhang Cai, Xiang Zhi, Jiwen Wu, Yongjun Xu, Qian Zhang

ECCV 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Knowledge distillation for detectors · 94 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Mixup task loss with auxiliary branches (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.08 -> 77.68Improved
add feature l2 Self-KD (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.68 -> 78.14Improved
add adversarial discriminator loss to feature Self-KD (ablation)MixSKD (ResNet-50) · ImageNettop-1: 78.14 -> 78.47Improved
add logit Self-KD over auxiliary branches (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.68 -> 78.06Improved
add self-teacher training and backbone logit Self-KD (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.68 -> 78.23Improved
combine all logit Self-KD losses (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.68 -> 78.54Improved
combine feature and logit Self-KD (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.68 -> 78.76Improved
build self-teacher from all stages instead of stage 1 only (ablation)MixSKD (ResNet-50) · ImageNettop-1: 78.01 -> 78.76Improved
build self-teacher from all stages instead of stage 4 only (ablation)MixSKD (ResNet-50) · ImageNettop-1: 78.42 -> 78.76Improved
set temperature T=3 instead of T=1 (ablation)MixSKD (ResNet-50) · ImageNettop-1: 77.94 -> 78.76Improved
set temperature T=3 instead of T=5 (ablation)MixSKD (ResNet-50) · ImageNettop-1: 78.37 -> 78.76Improved
add MixSKD to WRN-16-2 under FGSM attack eps=0.0001WRN-16-2 + Mixup · CIFAR-100top-1 under FGSM: 63.1 -> 67.7Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy