Tackling Long-Tailed Category Distribution Under Domain Shifts
This addresses a common real-world issue for machine learning applications where data is both imbalanced and subject to domain shifts, though it is incremental as it builds on existing long-tailed and domain generalization approaches.
The paper tackles the problem of long-tailed classification under domain shifts, where models fail due to both imbalanced training data and test data from different distributions, and demonstrates superior performance over state-of-the-art methods on new datasets AWA2-LTS and ImageNet-LTS.
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.