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

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 1–10 of 131 papers

Title	Date	Tasks	Status	Hype
Visual Prompt Tuning	Mar 23, 2022	Image ClassificationLong-tail Learning	CodeCode Available	3
Adaptive Parametric Activation	Jul 11, 2024	imbalanced classificationInstance Segmentation	CodeCode Available	2
Probabilistic Contrastive Learning for Long-Tailed Visual Recognition	Mar 11, 2024	Long-tail Learning	CodeCode Available	2
SURE: SUrvey REcipes for building reliable and robust deep networks	Mar 1, 2024	image-classificationImage Classification	CodeCode Available	2
Generalized Parametric Contrastive Learning	Sep 26, 2022	Contrastive LearningDomain Generalization	CodeCode Available	2
BatchFormer: Learning to Explore Sample Relationships for Robust Representation Learning	Mar 3, 2022	Compositional Zero-Shot LearningContrastive Learning	CodeCode Available	2
A Simple Episodic Linear Probe Improves Visual Recognition in the Wild	Jan 1, 2022	Fine-Grained Image ClassificationImage Classification	CodeCode Available	2
Learning Transferable Visual Models From Natural Language Supervision	Feb 26, 2021	Action RecognitionBenchmarking	CodeCode Available	2
Focal Loss for Dense Object Detection	Aug 7, 2017	2D Object DetectionDense Object Detection	CodeCode Available	2
Improving Visual Prompt Tuning by Gaussian Neighborhood Minimization for Long-Tailed Visual Recognition	Oct 28, 2024	Long-tail LearningVisual Prompt Tuning	CodeCode Available	1

Show:10 25 50

← PrevPage 1 of 14Next →

All datasets ImageNet-LT CIFAR-100-LT (ρ=100)CIFAR-10-LT (ρ=10)iNaturalist 2018 CIFAR-100-LT (ρ=10)Places-LT CIFAR-10-LT (ρ=100)CIFAR-100-LT (ρ=50)MIMIC-CXR-LT NIH-CXR-LT COCO-MLT VOC-MLT

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	LIFT (ViT-L/14)	Top-1 Accuracy	82.9	—	Unverified
2	µ2Net+ (ViT-L/16)	Top-1 Accuracy	82.5	—	Unverified
3	MAM (ViT-B/16)	Top-1 Accuracy	82.3	—	Unverified
4	LIFT (ViT-B/16)	Top-1 Accuracy	78.3	—	Unverified
5	VL-LTR (ViT-B-16)	Top-1 Accuracy	77.2	—	Unverified
6	BALLAD(ResNet-50×16)	Top-1 Accuracy	76.5	—	Unverified
7	BALLAD(ViT-B-16)	Top-1 Accuracy	75.7	—	Unverified
8	BALLAD(ResNet-101)	Top-1 Accuracy	70.5	—	Unverified
9	VL-LTR (ResNet-50)	Top-1 Accuracy	70.1	—	Unverified
10	BALLAD(ResNet-50)	Top-1 Accuracy	67.2	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	Cross-Entropy (CE)	Error Rate	62.75	—	Unverified
2	Cross-Entropy (CE)	Error Rate	61.68	—	Unverified
3	IBLLoss	Error Rate	61.52	—	Unverified
4	Cross-Entropy + Curvature Regularization	Error Rate	59.5	—	Unverified
5	CE-DRW	Error Rate	58.9	—	Unverified
6	LDAM-DRW	Error Rate	57.96	—	Unverified
7	ELP	Error Rate	57.6	—	Unverified
8	CDB-loss	Error Rate	57.43	—	Unverified
9	CE-DRW-IC	Error Rate	56.9	—	Unverified
10	LDAM-DRW + SSP	Error Rate	56.57	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	RISDA	Error Rate	20.11	—	Unverified
2	Empirical Risk Minimization (ERM, CE)	Error Rate	13.61	—	Unverified
3	Class-balanced Reweighting	Error Rate	13.46	—	Unverified
4	Class-balanced Resampling	Error Rate	13.21	—	Unverified
5	IBLLoss	Error Rate	12.93	—	Unverified
6	Class-balanced Focal Loss	Error Rate	12.9	—	Unverified
7	DecTDE	Error Rate	12.63	—	Unverified
8	M2m	Error Rate	12.5	—	Unverified
9	Prior-LT	Error Rate	12.2	—	Unverified
10	ELF&LDAM+DRW	Error Rate	12	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	LIFT (ViT-L/14@336px)	Top-1 Accuracy	87.4	—	Unverified
2	LIFT (ViT-L/14)	Top-1 Accuracy	85.2	—	Unverified
3	GML (ViT-B-16)	Top-1 Accuracy	82.1	—	Unverified
4	LIFT (ViT-B/16)	Top-1 Accuracy	80.4	—	Unverified
5	RAC (ViT-B-16)	Top-1 Accuracy	80.24	—	Unverified
6	GPaCo (2-R152)	Top-1 Accuracy	79.8	—	Unverified
7	TADE(ResNet-152)	Top-1 Accuracy	77	—	Unverified
8	ProCo (ResNet50)	Top-1 Accuracy	75.8	—	Unverified
9	MDCS(Resnet50)	Top-1 Accuracy	75.6	—	Unverified
10	DeiT-LT	Top-1 Accuracy	75.1	—	Unverified