Auditing Demographic Bias in Facial Landmark Detection for Fair Human-Robot Interaction

2026-04-08

AI Summary

This study investigates fairness biases in facial landmark detection arising from demographic attributes such as age, gender, and race. To address confounding factors—such as head pose and image resolution—that commonly interfere with fairness assessments in low-level vision tasks, the authors propose a statistical auditing framework capable of disentangling demographic attributes from visual covariates. Experimental results demonstrate that, after controlling for these confounders, performance disparities related to gender and race become statistically insignificant, whereas accuracy for older individuals remains substantially lower. This finding provides the first clear evidence of genuine age-related bias in facial landmark detection. The work introduces a novel methodology for evaluating fairness in low-level vision models and highlights their potential implications for human-computer interaction systems.

Abstract

Fairness in human-robot interaction critically depends on the reliability of the perceptual models that enable robots to interpret human behavior. While demographic biases have been widely studied in high-level facial analysis tasks, their presence in facial landmark detection remains unexplored. In this paper, we conduct a systematic audit of demographic bias in this task, analyzing the age, gender and race biases. To this end we introduce a controlled statistical methodology to disentangle demographic effects from confounding visual factors. Evaluations of a standard representative model demonstrate that confounding visual factors, particularly head pose and image resolution, heavily outweigh the impact of demographic attributes. Notably, after accounting for these confounders, we show that performance disparities across gender and race vanish. However, we identify a statistically significant age-related effect, with higher biases observed for older individuals. This shows that fairness issues can emerge even in low-level vision components and can propagate through the HRI pipeline, disproportionately affecting vulnerable populations. We argue that auditing and correcting such biases is a necessary step toward trustworthy and equitable robot perception systems.

Problem

• demographic bias
• facial landmark detection
• fairness
• human-robot interaction
• age bias

Innovation

• demographic bias
• facial landmark detection
• confounding factors
• fairness auditing
• human-robot interaction