Home Research Papers Datasets State of the Art Pricing

Discover, visualize, and connect AI research papers. Explore the latest trends and insights in artificial intelligence research.

Product

Home
Research Papers
About

Support

Contact
Terms of Service
Privacy Policy

© 2026 Papersgraph. All rights reserved.

methodology-8

Long-tail Learning

3260 papers • 126 benchmarks • 313 datasets

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.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in long-tail-learning-24

Trend

Dataset

Best Model

Actions

ImageNet-LT

ImageNet-LT

CIFAR-100-LT (ρ=100)

CIFAR-100-LT (ρ=100)

CIFAR-10-LT (ρ=10)

Libraries

i

Use these libraries to find long-tail-learning-24 models and implementations

Datasets

CIFAR-10

ImageNet

CIFAR-100

iNaturalist

ImageNet-LT

Subtasks

Long-tail learning with class descriptors

Most implemented papers

Learning Transferable Visual Models From Natural Language Supervision

I. Sutskever, Alec Radford, Jong Wook Kim, Chris Hallacy, A. Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger•Thu Feb 25 2021

It is demonstrated that the simple pre-training task of predicting which caption goes with which image is an efficient and scalable way to learn SOTA image representations from scratch on a dataset of 400 million (image, text) pairs collected from the internet.

39743

Content

Introduction Benchmarks Datasets Subtasks Libraries Papers

CIFAR-10-LT (ρ=10)

iNaturalist 2018

iNaturalist 2018

CIFAR-100-LT (ρ=10)

CIFAR-100-LT (ρ=10)

Places-LT

Places-LT

CIFAR-10-LT (ρ=100)

CIFAR-10-LT (ρ=100)

CIFAR-100-LT (ρ=50)

CIFAR-100-LT (ρ=50)

NIH-CXR-LT

NIH-CXR-LT

MIMIC-CXR-LT

MIMIC-CXR-LT

COCO-MLT

COCO-MLT

VOC-MLT

VOC-MLT

ImageNet-GLT

ImageNet-GLT

CIFAR-10-LT (ρ=50)

CIFAR-10-LT (ρ=50)

EGTEA

EGTEA

CIFAR-10-LT (ρ=200)

CIFAR-10-LT (ρ=200)

CIFAR-100-LT (ρ=200)

CIFAR-100-LT (ρ=200)

mini-ImageNet-LT

mini-ImageNet-LT

CelebA-5

CelebA-5

Lot-insts

Lot-insts

EGTEA

EGTEA

Places-LT

COCO-MLT

VOC-MLT

Animal Kingdom

0

Class-Balanced Loss Based on Effective Number of Samples

Yang Song, Yin Cui, Serge J. Belongie, Tsung-Yi Lin, Menglin Jia•Tue Jan 15 2019

This work designs a re-weighting scheme that uses the effective number of samples for each class to re-balance the loss, thereby yielding a class-balanced loss and introduces a novel theoretical framework to measure data overlap by associating with each sample a small neighboring region rather than a single point.

2787 0

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

Tengyu Ma, Adrien Gaidon, Colin Wei, Kaidi Cao, Nikos Aréchiga•Mon Jun 17 2019

A theoretically-principled label-distribution-aware margin (LDAM) loss motivated by minimizing a margin-based generalization bound is proposed that 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.

1918 0

ResLT: Residual Learning for Long-Tailed Recognition

Jiaya Jia, Shu Liu, Zhuotao Tian, Zhisheng Zhong, Jiequan Cui•Mon Jan 25 2021

This work designs the effective residual fusion mechanism – with one main branch optimized to recognize images from all classes, another two residual branches are gradually fused and optimized to enhance images from medium+tail classes and tail classes respectively.

183 0

Improving Calibration for Long-Tailed Recognition

Jiaya Jia, Shu Liu, Zhisheng Zhong, Jiequan Cui•Sun Sep 27 2020

Motivated by the fact that predicted probability distributions of classes are highly related to the numbers of class instances, this work proposes label-aware smoothing to deal with different degrees of over-confidence for classes and improve classifier learning.

377 0

Parametric Contrastive Learning

Jiaya Jia, Bei Yu, Shu Liu, Zhisheng Zhong, Jiequan Cui•Sun Jul 25 2021

This paper introduces a set of parametric class-wise learnable centers to rebalance from an optimization perspective and demonstrates that PaCo can adaptively enhance the intensity of pushing samples of the same class close as more samples are pulled together with their corresponding centers and benefit hard example learning.

355 0

Decoupling Representation and Classifier for Long-Tailed Recognition

Jiashi Feng, Bingyi Kang, Saining Xie, Marcus Rohrbach, Zhicheng Yan, Albert Gordo, Yannis Kalantidis•Sun Oct 20 2019

It is shown that it is possible to outperform carefully designed losses, sampling strategies, even complex modules with memory, by using a straightforward approach that decouples representation and classification.

1376 0

Generalized Parametric Contrastive Learning

Jiaya Jia, Bei Yu, Zhuotao Tian, Shu Liu, Zhisheng Zhong, Jiequan Cui•Sun Sep 25 2022

This analysis demonstrates that GPaCo/PaCo can adaptively enhance the intensity of pushing samples of the same class close as more samples are pulled together with their corresponding centers and benefit hard example learning.

70 0

Long-tail learning via logit adjustment

Sanjiv Kumar, Sadeep Jayasumana, Andreas Veit, A. Rawat, A. Menon, Himanshu Jain•Mon Jul 13 2020

These techniques revisit the classic idea of logit adjustment based on the label frequencies, either applied post-hoc to a trained model, or enforced in the loss during training, to encourage a large relative margin between logits of rare versus dominant labels.

877 0

Adding a benchmark result helps the community track progress.

Long-tail Learning | State-of-the-Art