H2T-FAST: Head-to-Tail Feature Augmentation by Style Transfer for Long-Tailed Recognition

Ziyao Meng,Xue Gu,Qiang Shen,Adriano J.C Tavares,Sandro Pinto,Hao Xu
2023-01-01
Abstract:Deep learning algorithms perform poorly on long-tailed datasets because there is insufficient data in the tail classes to recover its original distribution, resulting in an under-representation of the tail classes in the model. In this work, we propose H2T-FAST, a Head-to-Tail Feature Augmentation method by Style Transfer to improve the performance of the tail. H2T-FAST has the following advantages: (1) It is a fast and universal method that acts on the feature space and so, it can be applied to different backbone networks as well as easily integrated into various imbalanced algorithms with stable performance gains; and (2) it is used only in the training phase and therefore, imposes no additional burden on the deep neural network in the testing phase. In particular, we firstly and randomly select the same number of head samples as the tail ones in each training mini batch. Secondly, the style of the head is transferred to the tail to generate new tail data containing the head style, as a way to increase the number of the tail and get better feature representations. We test our methods on several benchmark vision tasks with state-of-the-art performances.
What problem does this paper attempt to address?