Fairness Without Labels: Pseudo-Balancing for Bias Mitigation in Face Gender Classification
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908587655168000 |
|---|---|
| author | Dong, Haohua Rodríguez, Ana Manzano Guinaudeau, Camille Satoh, Shin'ichi |
| author_facet | Dong, Haohua Rodríguez, Ana Manzano Guinaudeau, Camille Satoh, Shin'ichi |
| contents | Face gender classification models often reflect and amplify demographic biases present in their training data, leading to uneven performance across gender and racial subgroups. We introduce pseudo-balancing, a simple and effective strategy for mitigating such biases in semi-supervised learning. Our method enforces demographic balance during pseudo-label selection, using only unlabeled images from a race-balanced dataset without requiring access to ground-truth annotations.
We evaluate pseudo-balancing under two conditions: (1) fine-tuning a biased gender classifier using unlabeled images from the FairFace dataset, and (2) stress-testing the method with intentionally imbalanced training data to simulate controlled bias scenarios. In both cases, models are evaluated on the All-Age-Faces (AAF) benchmark, which contains a predominantly East Asian population. Our results show that pseudo-balancing consistently improves fairness while preserving or enhancing accuracy. The method achieves 79.81% overall accuracy - a 6.53% improvement over the baseline - and reduces the gender accuracy gap by 44.17%. In the East Asian subgroup, where baseline disparities exceeded 49%, the gap is narrowed to just 5.01%. These findings suggest that even in the absence of label supervision, access to a demographically balanced or moderately skewed unlabeled dataset can serve as a powerful resource for debiasing existing computer vision models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_10191 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Fairness Without Labels: Pseudo-Balancing for Bias Mitigation in Face Gender Classification Dong, Haohua Rodríguez, Ana Manzano Guinaudeau, Camille Satoh, Shin'ichi Computer Vision and Pattern Recognition 68T07 I.2.10; I.4.8; I.5.4 Face gender classification models often reflect and amplify demographic biases present in their training data, leading to uneven performance across gender and racial subgroups. We introduce pseudo-balancing, a simple and effective strategy for mitigating such biases in semi-supervised learning. Our method enforces demographic balance during pseudo-label selection, using only unlabeled images from a race-balanced dataset without requiring access to ground-truth annotations. We evaluate pseudo-balancing under two conditions: (1) fine-tuning a biased gender classifier using unlabeled images from the FairFace dataset, and (2) stress-testing the method with intentionally imbalanced training data to simulate controlled bias scenarios. In both cases, models are evaluated on the All-Age-Faces (AAF) benchmark, which contains a predominantly East Asian population. Our results show that pseudo-balancing consistently improves fairness while preserving or enhancing accuracy. The method achieves 79.81% overall accuracy - a 6.53% improvement over the baseline - and reduces the gender accuracy gap by 44.17%. In the East Asian subgroup, where baseline disparities exceeded 49%, the gap is narrowed to just 5.01%. These findings suggest that even in the absence of label supervision, access to a demographically balanced or moderately skewed unlabeled dataset can serve as a powerful resource for debiasing existing computer vision models. |
| title | Fairness Without Labels: Pseudo-Balancing for Bias Mitigation in Face Gender Classification |
| topic | Computer Vision and Pattern Recognition 68T07 I.2.10; I.4.8; I.5.4 |
| url | https://arxiv.org/abs/2510.10191 |