One Label, One Billion Faces: Usage and Consistency of Racial Categories in Computer Vision

Abstract

Computer vision is widely deployed, has highly visible, society altering applications, and documented problems with bias and representation. Datasets are critical for benchmarking progress in fair computer vision, and often employ broad racial categories as population groups for measuring group fairness. Similarly, diversity is often measured in computer vision datasets by ascribing and counting categorical race labels. However, racial categories are ill-defined, unstable temporally and geographically, and have a problematic history of scientific use. Although the racial categories used across datasets are superficially similar, the complexity of human race perception suggests the racial system encoded by one dataset may be substantially inconsistent with another. Using the insight that a classifier can learn the racial system encoded by a dataset, we conduct an empirical study of computer vision datasets supplying categorical race labels for face images to determine the cross-dataset consistency and generalization of racial categories. We find that each dataset encodes a substantially unique racial system, despite nominally equivalent racial categories, and some racial categories are systemically less consistent than others across datasets. We find evidence that racial categories encode stereotypes, and exclude ethnic groups from categories on the basis of nonconformity to stereotypes. Representing a billion humans under one racial category may obscure disparities and create new ones by encoding stereotypes of racial systems. The difficulty of adequately converting the abstract concept of race into a tool for measuring fairness underscores the need for a method more flexible and culturally aware than racial categories.

Published as a conference paper at the 4th ACM Conference on Fairness, Accountability, and Transparency (FAccT 2021)

References in corpus (7)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Understanding deep learning requires rethinking generalization
• fastai: A Layered API for Deep Learning
• Towards a Critical Race Methodology in Algorithmic Fairness
• Characterizing the Variability in Face Recognition Accuracy Relative to Race
• How Transferable are CNN-based Features for Age and Gender Classification?
• Generative One-Shot Face Recognition

Cited by in corpus (6)

• Gender and Racial Bias in Visual Question Answering Datasets
• Out of Context: Investigating the Bias and Fairness Concerns of "Artificial Intelligence as a Service"
• Skin Deep: Investigating Subjectivity in Skin Tone Annotations for Computer Vision Benchmark Datasets
• An Empirical Analysis of Racial Categories in the Algorithmic Fairness Literature
• Evaluating the Social Impact of Generative AI Systems in Systems and Society
• Algorithmic Fairness Datasets: the Story so Far