AI Paper Atlas

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

AREA: Adaptive Reweighting via Effective Area for Long-Tailed Classification

Xiaohua Chen, Yucan Zhou, Dayan Wu, Chule Yang, Bo Li, Qinghua Hu, Weiping Wang

ICCV 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Classifier rebalancing and margin losses · 71 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
replace sample-count weights with effective-area weightsCE-DRW / Class-Balanced Loss · CIFAR-100-LT (imbalance 100)accuracy (%): 41.51 -> 48.83 (vs CE-DRW)Improved
adaptively update weights during training instead of fixing after stage-I (ablation)AREA · CIFAR-100-LT (imbalance 100)accuracy (%): 47.73 -> 48.83Improved
adaptively update weights during training instead of fixing after stage-I (ablation)AREA · CIFAR-100-LT (imbalance 200)accuracy (%): 42.23 -> 43.85Improved
apply AREA reweighting on balanced datasetsCross-Entropy · CIFAR-100 (balanced)accuracy (%): 68.80 -> 69.70Improved
apply AREA reweighting on balanced CIFAR-10Cross-Entropy · CIFAR-10 (balanced)accuracy (%): 92.61 -> 92.84Improved
plug AREA weights into SOTA long-tail methodsGLMC, NCL, PaCo, BCL · CIFAR-100-LT (imbalance 100)accuracy (%): 57.11 -> 57.95 (GLMC)Improved
add AREA vs Decoupling on ImageNet-LT many-shot classes against CECross-Entropy · ImageNet-LT Many-shotaccuracy (%): drop of 3.11% on ManyHurt
use AREA instead of IB on CIFAR-10-LT imbalance 10IB / Meta-Class-Weight · CIFAR-10-LT (imbalance 10)accuracy (%): 88.85 vs 88.71No clear effect
AREA gains grow with imbalance ratioAREA vs IB · CIFAR-100-LTaccuracy (%): +3.64% at 10 to +6.69% at 100Improved

Later papers that compared with this method

1 later accepted paper reports Area as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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