AI Paper Atlas

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

Long-Tailed Recognition via Weight Balancing

Shaden Alshammari, Yu-Xiong Wang, Deva Ramanan, Shu Kong

CVPR 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Decoupled and calibrated long-tail training · 147 citations (OpenAlex)

Check 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
Tune weight decay during cross-entropy feature learning (ablation)CIFAR100-LT (IF=100), ResNet32 · CIFAR100-LT (IF=100)top-1 accuracy: 38.38 -> 46.08Improved
Add tau-normalization to the tuned-WD model (ablation)CIFAR100-LT (IF=100), ResNet32 · CIFAR100-LT (IF=100)top-1 accuracy: 46.08 -> 51.31Improved
Add L2 normalization to the tuned-WD model (ablation)CIFAR100-LT (IF=100), ResNet32 · CIFAR100-LT (IF=100)top-1 accuracy: 46.08 -> 49.60Improved
Retrain the classifier with CB loss only (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 46.08 -> 47.09Improved
Add classifier weight decay with CB loss (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 47.09 -> 52.42Improved
Add MaxNorm to classifier weight decay and CB loss (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 52.42 -> 53.35Improved
Use L2 normalization rather than classifier WD and MaxNorm with CB loss (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 53.35 -> 50.10Hurt
Fine-tune the last two layers instead of only the classifier (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 53.35 -> 53.55Improved
Fine-tune the last five layers rather than only the classifier (ablation)Tuned-WD model on CIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 53.35 -> 51.45Hurt
Use WD + MaxNorm weight balancing instead of the naive modelCIFAR100-LT (IF=100) · CIFAR100-LT (IF=100)top-1 accuracy: 38.38 -> 53.35Improved
Use weight balancing instead of the naive modelImageNet-LT · ImageNet-LT testtop-1 accuracy: 38.0 -> 53.9Improved
Use weight balancing instead of the naive modeliNaturalist2018 · iNaturalist2018top-1 accuracy: 46.1 -> 70.2Improved

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