Authors
- Yusuke Hirota
- Jerone Andrews
- Dora Zhao*
- Orestis Papakyriakopoulos*
- Apostolos Modas
- Yuta Nakashima*
- Alice Xiang
* External authors
Venue
- EMNLP 2024
Date
- 2024
Share
Resampled Datasets Are Not Enough: Mitigating Societal Bias Beyond Single Attributes
Yusuke Hirota
Dora Zhao*
Orestis Papakyriakopoulos*
Apostolos Modas
Yuta Nakashima*
* External authors
EMNLP 2024
2024
Abstract
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.
Related Publications
Despite extensive efforts to create fairer machine learning (ML) datasets, there remains a limited understanding of the practical aspects of dataset curation. Drawing from interviews with 30 ML dataset curators, we present a comprehensive taxonomy of the challenges and trade…
Vision-language models (VLMs) pre-trained on extensive datasets can inadvertently learn biases by correlating gender information with specific objects or scenarios. Current methods, which focus on modifying inputs and monitoring changes in the model's output probability scor…
Deep neural networks trained via empirical risk minimisation often exhibit significant performance disparities across groups, particularly when group and task labels are spuriously correlated (e.g., “grassy background” and “cows”). Existing bias mitigation methods that aim t…
JOIN US
Shape the Future of AI with Sony AI
We want to hear from those of you who have a strong desire
to shape the future of AI.
