Tackling Long-Tailed Category Distribution Under Domain Shifts

TL;DR

This paper introduces a novel meta-learning approach with three core modules to address long-tailed category distribution challenges under domain shifts, improving generalization on unseen domains.

Contribution

It proposes a new framework combining Distribution Calibrated Classification Loss, Visual-Semantic Mapping, and Semantic-Similarity Guided Augmentation for long-tailed domain generalization.

Findings

Outperforms state-of-the-art methods on new long-tailed domain datasets

Introduces two datasets: AWA2-LTS and ImageNet-LTS

Demonstrates significant improvement in domain generalization performance

Abstract

Machine learning models fail to perform well on real-world applications when 1) the category distribution P(Y) of the training dataset suffers from long-tailed distribution and 2) the test data is drawn from different conditional distributions P(X|Y). Existing approaches cannot handle the scenario where both issues exist, which however is common for real-world applications. In this study, we took a step forward and looked into the problem of long-tailed classification under domain shifts. We designed three novel core functional blocks including Distribution Calibrated Classification Loss, Visual-Semantic Mapping and Semantic-Similarity Guided Augmentation. Furthermore, we adopted a meta-learning framework which integrates these three blocks to improve domain generalization on unseen target domains. Two new datasets were proposed for this problem, named AWA2-LTS and ImageNet-LTS. We evaluated our method on the two datasets and extensive experimental results demonstrate that our proposed method can achieve superior performance over state-of-the-art long-tailed/domain generalization approaches and the combinations. Source codes and datasets can be found at our project page https://xiaogu.site/LTDS.