Is Visual Realism Enough? Evaluating Gait Biometric Fidelity in Generative AI Human Animation

2025-12-22

AI Summary

This study investigates whether generative AI (GenAI) human animation models preserve fine-grained spatiotemporal gait dynamics—essential for biometric identification—under high visual fidelity, addressing their identity-motion entanglement problem. We propose a dual-task evaluation framework comprising reference video reconstruction and cross-identity gait transfer, integrating biometric accuracy metrics with motion-appearance disentanglement analysis to systematically assess four state-of-the-art GenAI animation models. Our key finding is that current models rely predominantly on texture and other appearance cues—not temporal motion patterns—for gait recognition; recognition accuracy drops sharply after identity transfer, confirming failure to disentangle identity and motion representations. These results reveal insufficient identity fidelity of existing GenAI animations at the biometric level and expose the fundamental limitation of appearance-driven generation paradigms. The study provides critical empirical evidence for designing identifiable, controllable generative models.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Natural Language Processing: Generation

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Generative AI (GenAI) models have revolutionized animation, enabling the synthesis of humans and motion patterns with remarkable visual fidelity. However, generating truly realistic human animation remains a formidable challenge, where even minor inconsistencies can make a subject appear unnatural. This limitation is particularly critical when AI-generated videos are evaluated for behavioral biometrics, where subtle motion cues that define identity are easily lost or distorted. The present study investigates whether state-of-the-art GenAI human animation models can preserve the subtle spatio-temporal details needed for person identification through gait biometrics. Specifically, we evaluate four different GenAI models across two primary evaluation tasks to assess their ability to i) restore gait patterns from reference videos under varying conditions of complexity, and ii) transfer these gait patterns to different visual identities. Our results show that while visual quality is mostly high, biometric fidelity remains low in tasks focusing on identification, suggesting that current GenAI models struggle to disentangle identity from motion. Furthermore, through an identity transfer task, we expose a fundamental flaw in appearance-based gait recognition: when texture is disentangled from motion, identification collapses, proving current GenAI models rely on visual attributes rather than temporal dynamics.

Problem

• Evaluates gait biometric fidelity in AI-generated human animations
• Tests if AI models preserve spatio-temporal details for person identification
• Examines identity transfer and motion-texture disentanglement in gait recognition

Innovation

• Evaluates gait biometric fidelity in AI animation
• Tests models on gait restoration and identity transfer
• Finds models rely on visual attributes over temporal dynamics