AI Paper Atlas

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

ImbSAM: A Closer Look at Sharpness-Aware Minimization in Class-Imbalanced Recognition

Yixuan Zhou, Yi Qu, Xing Xu, Hengtao Shen

ICCV 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Decoupled and calibrated long-tail training · 27 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
apply class-aware ImbSAM instead of class-agnostic SAMCE · CIFAR100-LT (IF=100)Few accuracy (%): 12.4 -> 19.4Improved
apply class-aware ImbSAM instead of class-agnostic SAMLTR · CIFAR100-LT (IF=100)Few accuracy (%): 25.3 -> 39.4Improved
add standard SAM to CECE · CIFAR100-LT (IF=100)Few accuracy (%): 13.5 -> 12.4Hurt
add large tail loss re-weighting (up to 20) to SAM (ablation)CE+SAM · CIFAR100-LT (IF=100)All accuracy (%): 48.9 -> 49.1No clear effect
apply ImbSAM instead of CE aloneCE · CIFAR100-LT (IF=100)Many accuracy (%): 77.8 -> 75.9Hurt
apply ImbSAM on top of WB (LTR)LTR · CIFAR100-LT (IF=100)All accuracy (%): 51.9 -> 54.8Improved
set class split threshold eta to the Few-class limit (20) (ablation)ImbSAM · CIFAR100-LT (IF=100)Few accuracy (%): up to 21.5%Improved
increase rho with eta=20 (ablation)ImbSAM · CIFAR100-LT (IF=100)Few accuracy (%)Hurt
apply standard SAM on large-scale long-tailed dataCE · ImageNet-LT and iNaturalistaccuracy on Many classesNo clear effect
apply ImbSAM to DeepSADDeepSAD · SSAD average of 5 datasetsAUCROC (%): 90.6 -> 91.8Improved
apply standard SAM to DeepSAD and DevNetDeepSAD, DevNet · SSAD average of 5 datasetsAUCROC (%): 90.6 -> 90.4; 91.6 -> 91.0Hurt
apply ImbSAM to DevNet at low anomaly ratioDevNet · SSAD AUCPR anomaly (gamma_l<0.1)AUCPR anomalyHurt

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