Identifying Ethical Biases in Action Recognition Models

arXiv:2604.17971v1 Announce Type: new Abstract: Human Action Recognition (HAR) models are increasingly deployed in high-stakes environments, yet their fairness across different human appearances has not been analyzed. We introduce a framework for auditing bias in HAR models using synthetic video data, generated with full control over visual identity attributes such as skin color. Unlike prior work that focuses on static images or pose estimation, our approach preserves temporal consistency, allowing us to isolate and test how changes to a single attribute affect model predictions. Through controlled interventions using the BEDLAM simulation platform, we show whether some popular HAR models exhibit statistically significant biases on the skin color even when the motion remains identical. Our results highlight how models may encode unwanted visual associations, and we provide evidence of systematic errors across groups. This work contributes a framework for auditing HAR models and supports the development of more transparent, accountable systems in light of upcoming regulatory standards.

Related

• When Fairness Metrics Disagree: Evaluating the Reliability of Demographic Fairness Assessment in Machine Learning
• Operationalizing Fairness in Text-to-Image Models: A Survey of Bias, Fairness Audits and Mitigation Strategies
• Bias at the End of the Score
• Bias Redistribution in Visual Machine Unlearning: Does Forgetting One Group Harm Another?
• Motion-Guided Semantic Alignment with Negative Prompts for Zero-Shot Video Action Recognition

Source: arXiv cs.CV | 2026-04-21