TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Long-tail learning with class descriptors	AWA-LT	LDAM	Per-Class Accuracy	69.1	# 5
Long-tail learning with class descriptors	AWA-LT	LDAM	Long-Tailed Accuracy	93.5	# 2
Long-tail Learning	CIFAR-100-LT (ρ=10)	LDAM-DRW	Error Rate	41.29	# 30
Long-tail Learning	CIFAR-100-LT (ρ=100)	LDAM-DRW	Error Rate	57.96	# 59
Long-tail Learning	CIFAR-10-LT (ρ=10)	Empirical Risk Minimization (ERM, CE)	Error Rate	13.61	# 49
Long-tail Learning	CIFAR-10-LT (ρ=10)	LDAM-DRW	Error Rate	11.84	# 38
Long-tail Learning	CIFAR-10-LT (ρ=10)	Class-balanced Resampling	Error Rate	13.21	# 47
Long-tail Learning	CIFAR-10-LT (ρ=100)	LDAM-DRW	Error Rate	22.97	# 25
Long-tail Learning	COCO-MLT	LDAM(ResNet-50)	Average mAP	40.53	# 13
Long-tail learning with class descriptors	CUB-LT	LDAM	Per-Class Accuracy	50.1	# 4
Long-tail learning with class descriptors	CUB-LT	LDAM	Long-Tailed Accuracy	64.1	# 4
Long-tail learning with class descriptors	SUN-LT	LDAM	Per-Class Accuracy	29.8	# 5
Long-tail learning with class descriptors	SUN-LT	LDAM	Long-Tailed Accuracy	36.4	# 4
Long-tail Learning	VOC-MLT	LDAM(ResNet-50)	Average mAP	70.73	# 12

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-with-class-descriptors-on)](https://paperswithcode.com/sota/long-tail-learning-with-class-descriptors-on?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-with-class-descriptors-on-2)](https://paperswithcode.com/sota/long-tail-learning-with-class-descriptors-on-2?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-with-class-descriptors-on-1)](https://paperswithcode.com/sota/long-tail-learning-with-class-descriptors-on-1?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-voc-mlt)](https://paperswithcode.com/sota/long-tail-learning-on-voc-mlt?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-coco-mlt)](https://paperswithcode.com/sota/long-tail-learning-on-coco-mlt?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-cifar-10-lt-r-100)](https://paperswithcode.com/sota/long-tail-learning-on-cifar-10-lt-r-100?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-cifar-100-lt-r-10)](https://paperswithcode.com/sota/long-tail-learning-on-cifar-100-lt-r-10?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-cifar-10-lt-r-10)](https://paperswithcode.com/sota/long-tail-learning-on-cifar-10-lt-r-10?p=learning-imbalanced-datasets-with-label)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/learning-imbalanced-datasets-with-label/long-tail-learning-on-cifar-100-lt-r-100)](https://paperswithcode.com/sota/long-tail-learning-on-cifar-100-lt-r-100?p=learning-imbalanced-datasets-with-label)`

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

NeurIPS 2019 · Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, Tengyu Ma ·

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes. We design two novel methods to improve performance in such scenarios. First, we propose a theoretically-principled label-distribution-aware margin (LDAM) loss motivated by minimizing a margin-based generalization bound. This loss replaces the standard cross-entropy objective during training and can be applied with prior strategies for training with class-imbalance such as re-weighting or re-sampling. Second, we propose a simple, yet effective, training schedule that defers re-weighting until after the initial stage, allowing the model to learn an initial representation while avoiding some of the complications associated with re-weighting or re-sampling. We test our methods on several benchmark vision tasks including the real-world imbalanced dataset iNaturalist 2018. Our experiments show that either of these methods alone can already improve over existing techniques and their combination achieves even better performance gains.

PDF Abstract NeurIPS 2019 PDF NeurIPS 2019 Abstract

Code

Add Remove Mark official

kaidic/LDAM-DRW official

627

j3soon/arxiv-utils

267

feidfoe/AdjustBnd4Imbalance

orparask/VS-Loss

jackhck/subclass-balancing-contrast…

See all 7 implementations

Tasks

Add Remove

Long-tail Learning

Long-tail learning with class descriptors

Datasets

Introduced in the Paper:

CIFAR100-LT

Used in the Paper:

CIFAR-10

CIFAR-100

CUB-200-2011

IMDb Movie Reviews

Tiny ImageNet

iNaturalist

AwA SUN VOC-MLT COCO-MLT

Results from the Paper

Edit

Ranked #4 on Long-tail learning with class descriptors on CUB-LT

Get a GitHub badge

Task	Dataset	Model	Metric Name	Metric Value	Global Rank	Benchmark
Long-tail learning with class descriptors	AWA-LT	LDAM	Per-Class Accuracy	69.1	# 5	Compare
Long-tail learning with class descriptors	AWA-LT	LDAM	Long-Tailed Accuracy	93.5	# 2	Compare
Long-tail Learning	CIFAR-100-LT (ρ=10)	LDAM-DRW	Error Rate	41.29	# 30	Compare
Long-tail Learning	CIFAR-100-LT (ρ=100)	LDAM-DRW	Error Rate	57.96	# 59	Compare
Long-tail Learning	CIFAR-10-LT (ρ=10)	Empirical Risk Minimization (ERM, CE)	Error Rate	13.61	# 49	Compare
Long-tail Learning	CIFAR-10-LT (ρ=10)	LDAM-DRW	Error Rate	11.84	# 38	Compare
Long-tail Learning	CIFAR-10-LT (ρ=10)	Class-balanced Resampling	Error Rate	13.21	# 47	Compare
Long-tail Learning	CIFAR-10-LT (ρ=100)	LDAM-DRW	Error Rate	22.97	# 25	Compare
Long-tail Learning	COCO-MLT	LDAM(ResNet-50)	Average mAP	40.53	# 13	Compare
Long-tail learning with class descriptors	CUB-LT	LDAM	Per-Class Accuracy	50.1	# 4	Compare
Long-tail learning with class descriptors	CUB-LT	LDAM	Long-Tailed Accuracy	64.1	# 4	Compare
Long-tail learning with class descriptors	SUN-LT	LDAM	Per-Class Accuracy	29.8	# 5	Compare
Long-tail learning with class descriptors	SUN-LT	LDAM	Long-Tailed Accuracy	36.4	# 4	Compare
Long-tail Learning	VOC-MLT	LDAM(ResNet-50)	Average mAP	70.73	# 12	Compare

Methods

Add Remove

No methods listed for this paper. Add relevant methods here

Edit Social Preview

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

Code Edit Add Remove Mark official

Tasks Edit Add Remove

Datasets Edit

Results from the Paper Edit

Methods Edit Add Remove

Code

Add Remove Mark official

Tasks

Add Remove

Datasets

Results from the Paper

Edit

Methods

Add Remove