Publications

BibTex format

@inproceedings{Gu:2022:10.1007/978-3-031-20050-2_42,
author = {Gu, X and Guo, Y and Li, Z and Qiu, J and Dou, Q and Liu, Y and Lo, B and Yang, G-Z},
doi = {10.1007/978-3-031-20050-2_42},
pages = {727--743},
publisher = {Springer},
title = {Tackling long-tailed category distribution under domain shifts},
url = {http://dx.doi.org/10.1007/978-3-031-20050-2_42},
year = {2022}
}

Download

RIS format (EndNote, RefMan)

TY  - CPAPER
AB  - 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.
AU  - Gu,X
AU  - Guo,Y
AU  - Li,Z
AU  - Qiu,J
AU  - Dou,Q
AU  - Liu,Y
AU  - Lo,B
AU  - Yang,G-Z
DO  - 10.1007/978-3-031-20050-2_42
EP  - 743
PB  - Springer
PY  - 2022///
SN  - 0302-9743
SP  - 727
TI  - Tackling long-tailed category distribution under domain shifts
UR  - http://dx.doi.org/10.1007/978-3-031-20050-2_42
UR  - https://link.springer.com/chapter/10.1007/978-3-031-20050-2_42
UR  - http://hdl.handle.net/10044/1/98523
ER  -

Download

DrBennyLo

Contact

Location

Summary

Citation

BibTex format

RIS format (EndNote, RefMan)