TAET: Two-Stage Adversarial Equalization Training on Long-Tailed Distributions
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915207407730688 |
|---|---|
| author | YuHang, Wang Guo, Junkang Liu, Aolei Wang, Kaihao Wu, Zaitong Liu, Zhenyu Yin, Wenfei Liu, Jian |
| author_facet | YuHang, Wang Guo, Junkang Liu, Aolei Wang, Kaihao Wu, Zaitong Liu, Zhenyu Yin, Wenfei Liu, Jian |
| contents | Adversarial robustness is a critical challenge in deploying deep neural networks for real-world applications. While adversarial training is a widely recognized defense strategy, most existing studies focus on balanced datasets, overlooking the prevalence of long-tailed distributions in real-world data, which significantly complicates robustness. This paper provides a comprehensive analysis of adversarial training under long-tailed distributions and identifies limitations in the current state-of-the-art method, AT-BSL, in achieving robust performance under such conditions. To address these challenges, we propose a novel training framework, TAET, which integrates an initial stabilization phase followed by a stratified equalization adversarial training phase. Additionally, prior work on long-tailed robustness has largely ignored the crucial evaluation metric of balanced accuracy. To bridge this gap, we introduce the concept of balanced robustness, a comprehensive metric tailored for assessing robustness under long-tailed distributions. Extensive experiments demonstrate that our method surpasses existing advanced defenses, achieving significant improvements in both memory and computational efficiency. This work represents a substantial advancement in addressing robustness challenges in real-world applications. Our code is available at: https://github.com/BuhuiOK/TAET-Two-Stage-Adversarial-Equalization-Training-on-Long-Tailed-Distributions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_01924 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TAET: Two-Stage Adversarial Equalization Training on Long-Tailed Distributions YuHang, Wang Guo, Junkang Liu, Aolei Wang, Kaihao Wu, Zaitong Liu, Zhenyu Yin, Wenfei Liu, Jian Machine Learning Artificial Intelligence Adversarial robustness is a critical challenge in deploying deep neural networks for real-world applications. While adversarial training is a widely recognized defense strategy, most existing studies focus on balanced datasets, overlooking the prevalence of long-tailed distributions in real-world data, which significantly complicates robustness. This paper provides a comprehensive analysis of adversarial training under long-tailed distributions and identifies limitations in the current state-of-the-art method, AT-BSL, in achieving robust performance under such conditions. To address these challenges, we propose a novel training framework, TAET, which integrates an initial stabilization phase followed by a stratified equalization adversarial training phase. Additionally, prior work on long-tailed robustness has largely ignored the crucial evaluation metric of balanced accuracy. To bridge this gap, we introduce the concept of balanced robustness, a comprehensive metric tailored for assessing robustness under long-tailed distributions. Extensive experiments demonstrate that our method surpasses existing advanced defenses, achieving significant improvements in both memory and computational efficiency. This work represents a substantial advancement in addressing robustness challenges in real-world applications. Our code is available at: https://github.com/BuhuiOK/TAET-Two-Stage-Adversarial-Equalization-Training-on-Long-Tailed-Distributions. |
| title | TAET: Two-Stage Adversarial Equalization Training on Long-Tailed Distributions |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2503.01924 |