IEEE Access | 卷:9 |
Implicit Stereotypes in Pre-Trained Classifiers | |
Nassim Dehouche1  | |
[1] Mahidol University International College, Salaya, Thailand; | |
关键词: Algorithmic fairness; facial recognition; deep learning; zero-shot classification; CLIP; | |
DOI : 10.1109/ACCESS.2021.3136898 | |
来源: DOAJ |
【 摘 要 】
Pre-trained deep learning models underpin many public-facing applications, and their propensity to reproduce implicit racial and gender stereotypes is an increasing source of concern. The risk of large-scale, unfair outcomes resulting from their use thus raises the need for technical tools to test and audit these systems. In this work, a dataset of 10,000 portrait photographs was generated and classified, using CLIP (Contrastive Language–Image Pretraining), according to six pairs of opposing labels describing a subject’s gender, ethnicity, attractiveness, friendliness, wealth, and intelligence. Label correlation was analyzed and significant associations, corresponding to common implicit stereotypes in culture and society, were found at the 99% significance level. A strong positive correlation was notably found between labels
【 授权许可】
Unknown