GaitProtector: Impersonation-Driven Gait De-Identification via Training-Free Diffusion Latent Optimization

TL;DR

GaitProtector is a training-free, diffusion-based gait de-identification framework that balances privacy and utility by adversarially optimizing gait silhouettes against pretrained diffusion priors.

Contribution

It introduces a novel, training-free diffusion latent optimization method for gait de-identification that leverages pretrained 3D diffusion models without retraining.

Findings

Achieves 56.7% impersonation success rate against gait recognition.

Reduces Rank-1 identification accuracy from 89.6% to 15.0%.

Maintains diagnostic utility with only a 17.2% decrease in accuracy.

Abstract

Conventional gait de-identification methods often encounter an inherent trade-off: they either provide insufficient identity suppression or introduce spatiotemporal distortions that impede structure-sensitive downstream applications. We propose GaitProtector, an impersonation-driven gait de-identification framework that formulates privacy protection as a unified objective with two tightly coupled components: (i) obfuscation, which repels the protected gait from the source identity, and (ii) impersonation, which attracts it toward a selected target identity. The target identity serves as a semantic anchor that biases optimization toward structurally plausible gait patterns under the pretrained diffusion prior, helping preserve dominant body shape and motion dynamics. We instantiate this idea through a training-free diffusion latent optimization pipeline. Instead of retraining a generator for each dataset, we invert each input silhouette sequence into the latent trajectory of a pretrained 3D video diffusion model and iteratively optimize latent codes with a differentiable adversarial objective to synthesize protected gaits. Experiments on the CASIA-B dataset show that GaitProtector achieves a 56.7% impersonation success rate under black-box gait recognition and reduces Rank-1 identification accuracy from 89.6% to 15.0%, while maintaining favorable visual and temporal quality. We further evaluate downstream utility on the Scoliosis1K dataset, where diagnostic accuracy decreases only from 91.4% to 74.2%. To the best of our knowledge, this work is the first to leverage pretrained 3D diffusion priors in a training-free manner for silhouette-based gait de-identification.