Improving Worst-Group Accuracy with a Filtering-Based Method

Citations

WEB OF SCIENCE

0
Citations

SCOPUS

0

초록

Spurious correlations in training data often cause models to underperform on minority groups. Existing methods that improve worst-group accuracy (WGA) typically rely on group annotations, complex optimization objectives or multi-stage pipelines, limiting their practical application.We make a key observation: robust classifiers utilize a sparse set of features, suggesting they selectively focus on core, task-essential features. Based on this insight, we propose a single-stage, annotation-free method that co-trains a learnable binary mask to structurally filter input features. End-to-end training is enabled via a Straight-Through Estimator (STE) and a Gradient Adjustment technique, with the mask learned solely from the main task loss. While many de-biasing methods improve WGA, they often overlook the critical trade-off with overall accuracy. We therefore propose a novel Trade-off Ratio (ToR) metric, which quantifies the efficiency of WGA gains relative to any loss in overall accuracy. On four benchmarks (Waterbirds, CelebA, CivilComments, MultiNLI), our method achieves state-of-the-art WGA on CivilComments of 81.9% on CivilComments and the highest ToR score of 12.8 on Waterbirds.

키워드

classification; de-biasing; group robustness; spurious correlation
제목
Improving Worst-Group Accuracy with a Filtering-Based Method
저자
Kim, Mingi; Ryu, Jongwon; Kim, Junyeong
DOI
10.1109/ACCESS.2025.3643350
발행일
2025
유형
Article
저널명
IEEE Access
권
13
페이지
214201 ~ 214210

파일 다운로드

Thumbnail