SOTAVerified

Long-tail Learning

Long-tailed learning, one of the most challenging problems in visual recognition, aims to train well-performing models from a large number of images that follow a long-tailed class distribution.

Papers

Showing 110 of 131 papers

TitleStatusHype
Mitigating Spurious Correlations with Causal Logit Perturbation0
LIFT+: Lightweight Fine-Tuning for Long-Tail LearningCode0
Improving Visual Prompt Tuning by Gaussian Neighborhood Minimization for Long-Tailed Visual RecognitionCode1
Learning from Neighbors: Category Extrapolation for Long-Tail Learning0
Continuous Contrastive Learning for Long-Tailed Semi-Supervised RecognitionCode1
AUCSeg: AUC-oriented Pixel-level Long-tail Semantic SegmentationCode1
Representation Norm Amplification for Out-of-Distribution Detection in Long-Tail LearningCode0
LTRL: Boosting Long-tail Recognition via Reflective LearningCode1
On Characterizing and Mitigating Imbalances in Multi-Instance Partial Label Learning0
Adaptive Parametric ActivationCode2
Show:102550
← PrevPage 1 of 14Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1LMPT(ViT-B/16)Average mAP66.19Unverified
2CLIP(ViT-B/16)Average mAP60.17Unverified
3LMPT(ResNet-50)Average mAP58.97Unverified
4LTML(ResNet-50)Average mAP56.9Unverified
5CLIP(ResNet-50)Average mAP56.19Unverified
6PG Loss(ResNet-50)Average mAP54.43Unverified
7DB Focal(ResNet-50)Average mAP53.55Unverified
8Focal Loss(ResNet-50)Average mAP49.46Unverified
9CB Loss(ResNet-50)Average mAP49.06Unverified
10RS(ResNet-50)Average mAP46.97Unverified