English

Extracting Clean and Balanced Subset for Noisy Long-tailed Classification

Machine Learning 2024-04-11 v1

Abstract

Real-world datasets usually are class-imbalanced and corrupted by label noise. To solve the joint issue of long-tailed distribution and label noise, most previous works usually aim to design a noise detector to distinguish the noisy and clean samples. Despite their effectiveness, they may be limited in handling the joint issue effectively in a unified way. In this work, we develop a novel pseudo labeling method using class prototypes from the perspective of distribution matching, which can be solved with optimal transport (OT). By setting a manually-specific probability measure and using a learned transport plan to pseudo-label the training samples, the proposed method can reduce the side-effects of noisy and long-tailed data simultaneously. Then we introduce a simple yet effective filter criteria by combining the observed labels and pseudo labels to obtain a more balanced and less noisy subset for a robust model training. Extensive experiments demonstrate that our method can extract this class-balanced subset with clean labels, which brings effective performance gains for long-tailed classification with label noise.

Keywords

Cite

@article{arxiv.2404.06795,
  title  = {Extracting Clean and Balanced Subset for Noisy Long-tailed Classification},
  author = {Zhuo Li and He Zhao and Zhen Li and Tongliang Liu and Dandan Guo and Xiang Wan},
  journal= {arXiv preprint arXiv:2404.06795},
  year   = {2024}
}
R2 v1 2026-06-28T15:49:36.690Z