Gao Peifeng — accepted AI-conference papers

2023

Feature Directions Matter: Long-Tailed Learning via Rotated Balanced Representation

Gao Peifeng, Qianqian Xu, Peisong Wen, Zhiyong Yang, Huiyang Shao, Qingming Huang

ICML 2023poster

Long-tailed learning is one of the most challenging problems in visual recognition. There are some studies aiming to solve long-tailed classification from the perspective of feature learning. Recent work proposes to learn the balanced representation by fixing the linear classifier as Equiangular Tig…

Cited by 19 Source ↗PDF ↗Scholar ↗

2023

Weighted ROC Curve in Cost Space: Extending AUC to Cost-Sensitive Learning

Huiyang Shao, Qianqian Xu, Zhiyong Yang, Peisong Wen, Gao Peifeng, Qingming Huang

NeurIPS 2023poster

In this paper, we aim to tackle flexible cost requirements for long-tail datasets, where we need to construct a (a) cost-sensitive and (b) class-distribution robust learning framework. The misclassification cost and the area under the ROC curve (AUC) are popular metrics for (a) and (b), respectively…

Cited by 9 Source ↗PDF ↗Scholar ↗