Balanced Gradient Penalty Improves Deep Long-Tailed Learning

Dong Wang, Yicheng Liu, Liangji Fang, Fanhua Shang, Yuanyuan Liu, Hongying Liu

Abstract

In recent years, deep learning has achieved a great success in various image recognition tasks. However, the long-tailed setting over a semantic class plays a leading role in real-world applications. Common methods focus on optimization on balanced distribution or naive models. Few works explore long-tailed learning from a deep learning-based generalization perspective. The loss landscape on long-tailed learning is first investigated in this work. Empirical results show that sharpness-aware optimizers work not well on long-tailed learning. Because they do not take class priors into consideration, and they fail to improve performance of few-shot classes. To better guide the network and explicitly alleviate sharpness without extra computational burden, we develop a universal Balanced Gradient Penalty (BGP) method. Surprisingly, our BGP method does not need the detailed class priors and preserves privacy. Our new algorithm BGP, as a regularization loss, can achieve the state-of-the-art results on various image datasets (i.e., CIFAR-LT, ImageNet-LT and iNaturalist-2018) in the settings of different imbalance ratios.
BibTeX
@inproceedings{wang2022balanced,
  author = {Dong Wang and Yicheng Liu and Liangji Fang and Fanhua Shang and Yuanyuan Liu and Hongying Liu},
  title = {{Balanced Gradient Penalty Improves Deep Long-Tailed Learning}},
  booktitle = {ACM Multimedia},
  year = {2022},
  doi = {10.1145/3503161.3547763},
}

Loading related content...

Figure