AI Paper Atlas

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

Procrustean Training for Imbalanced Deep Learning

Han-Jia Ye, De-Chuan Zhan, Wei-Lun Chao

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 30 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 Major Feature WeakeningERM · CIFAR-10 step (rho=100)accuracy: 65.8 -> 80.1Improved
add Major Feature WeakeningERM · CIFAR-10 long-tailed (rho=100)accuracy: 71.1 -> 78.5Improved
add Deferred Re-Weight to MFW (ablation)MFW · CIFAR-10 step (rho=100)accuracy: 80.1 -> 81.6Improved
add DRW to MFW (ablation)MFW · CIFAR-100 step (rho=100)accuracy: 46.9 -> 48.4Improved
add DRW to MFW (ablation)MFW · CIFAR-100 (rho=10) long-tailedaccuracy: 60.1 -> 59.1Hurt
apply MFW after 2nd conv group rather than input layer (ablation)MFW · CIFAR-10 step (rho=100)accuracy: 68.1 -> 80.1Improved
apply MFW after 2nd group rather than 4th group (ablation)MFW · CIFAR-100 long-tailed (rho=100)accuracy: 40.3 -> 44.7Improved
keep labels unmixed (MFW) instead of mixup (ablation)mixup · CIFAR-10 long-tailed (rho=100)accuracy: 73.1 -> 78.5Improved
increase beta-distribution alpha from 0.1 to 5 (ablation)MFW · CIFAR-10 stepaccuracy: 73.3 -> 80.1Improved
increase alpha from 0.1 to 1 (ablation)MFW · CIFAR-100 long-tailedaccuracy: 43.1 -> 42.1Mixed
add MFW under extreme imbalance rho=1000ERM · CIFAR-10 long-tailed (rho=1000)accuracy: 56.5 -> 62.0Improved
add MFW to a DenseNet-121 backboneDenseNet-121 · CIFAR-10 long-tailed (rho=100)accuracy: 77.6 -> 81.2Improved

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