Apostolos Modas
2024
Resampled Datasets Are Not Enough: Mitigating Societal Bias Beyond Single Attributes
Yusuke Hirota
|
Jerone Andrews
|
Dora Zhao
|
Orestis Papakyriakopoulos
|
Apostolos Modas
|
Yuta Nakashima
|
Alice Xiang
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
We tackle societal bias in image-text datasets by removing spurious correlations between protected groups and image attributes. Traditional methods only target labeled attributes, ignoring biases from unlabeled ones. Using text-guided inpainting models, our approach ensures protected group independence from all attributes and mitigates inpainting biases through data filtering. Evaluations on multi-label image classification and image captioning tasks show our method effectively reduces bias without compromising performance across various models. Specifically, we achieve an average societal bias reduction of 46.1% in leakage-based bias metrics for multi-label classification and 74.8% for image captioning.
Search
Co-authors
- Yusuke Hirota 1
- Jerone Andrews 1
- Dora Zhao 1
- Orestis Papakyriakopoulos 1
- Yuta Nakashima 1
- show all...