AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

AutoMix: Unveiling the Power of Mixup for Stronger Classifiers

Zicheng Liu, Siyuan Li, Di Wu, Zihan Liu, Zhiyuan Chen, Lirong Wu, Stan Z. Li

ECCV 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 89 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 cross-attention, lambda embedding and lambda loss to MixBlock (ablation)AutoMix · Tiny-ImageNettop-1 accuracy: 66.83 -> 70.72Improved
add EMA and CE to the momentum pipeline of MixBlock (ablation)AutoMix · CIFAR-100 and Tiny-ImageNettop-1 accuracyImproved
MixBlock without EMA and CE versus vanilla training (ablation)MixBlock · CIFAR-100top-1 accuracy: +2.29%Improved
MixBlock without EMA and CE versus vanilla training (ablation)MixBlock · Tiny-ImageNettop-1 accuracy: +2.21%Improved
add CE loss to other mixup methods (ablation)other mixup methods · CIFAR-100top-1 accuracyNo clear effect
set mixup alpha to 2 instead of other values (ablation)AutoMix · CIFAR-100 and Tiny-ImageNettop-1 accuracyImproved
use feature layer l=3 for mixing (ablation)AutoMix · Tiny-ImageNettop-1 accuracyImproved
use AutoMix instead of AutoMix-free ResNet-18 baseline on ImageNet 100 epochs (ablation)ResNet-18 · ImageNet-1ktop-1 accuracy: 70.04 -> 70.50Improved
use AutoMix instead of vanilla on EfficientNet B0EfficientNet B0 · ImageNet-1ktop-1 accuracy: +0.32%Improved
use CutMix with longer training on CIFAR-100CutMix · CIFAR-100top-1 accuracy: 79.58 -> 78.29Hurt
use AutoMix instead of PuzzleMix on Tiny-ImageNet ResNeXt-50ResNeXt-50 · Tiny-ImageNettop-1 accuracy: 67.83 -> 70.72Improved
use AutoMix instead of vanilla on CUB-200ResNet-18 · CUB-200top-1 accuracy: 77.68 -> 79.87Improved

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