Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2404.06795 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910404923359232 |
|---|---|
| author | Li, Zhuo Zhao, He Li, Zhen Liu, Tongliang Guo, Dandan Wan, Xiang |
| author_facet | Li, Zhuo Zhao, He Li, Zhen Liu, Tongliang Guo, Dandan Wan, Xiang |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_06795 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Extracting Clean and Balanced Subset for Noisy Long-tailed Classification Li, Zhuo Zhao, He Li, Zhen Liu, Tongliang Guo, Dandan Wan, Xiang Machine Learning 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. |
| title | Extracting Clean and Balanced Subset for Noisy Long-tailed Classification |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2404.06795 |