Long-tailed Distribution Adaptation

TL;DR

This paper introduces a novel domain adaptation framework for long-tailed recognition, modeling the problem as unbalanced-to-balanced domain transfer, and achieves state-of-the-art results with interpretability.

Contribution

It formulates long-tailed recognition as domain adaptation, proposing a method that jointly optimizes risks and domain divergence to improve model adaptation.

Findings

Achieves state-of-the-art performance on benchmark datasets.

Provides an interpretable approach to long-tailed recognition.

Validates effectiveness across image recognition, detection, and segmentation.

Abstract

Recognizing images with long-tailed distributions remains a challenging problem while there lacks an interpretable mechanism to solve this problem. In this study, we formulate Long-tailed recognition as Domain Adaption (LDA), by modeling the long-tailed distribution as an unbalanced domain and the general distribution as a balanced domain. Within the balanced domain, we propose to slack the generalization error bound, which is defined upon the empirical risks of unbalanced and balanced domains and the divergence between them. We propose to jointly optimize empirical risks of the unbalanced and balanced domains and approximate their domain divergence by intra-class and inter-class distances, with the aim to adapt models trained on the long-tailed distribution to general distributions in an interpretable way. Experiments on benchmark datasets for image recognition, object detection, and instance segmentation validate that our LDA approach, beyond its interpretability, achieves state-of-the-art performance. Code is available at https://github.com/pengzhiliang/LDA.