PixelFade: Privacy-preserving Person Re-identification with Noise-guided Progressive Replacement

Delong Zhang, Yi-Xing Peng, Xiao-Ming Wu, Ancong Wu, Wei-Shi Zheng

Abstract: Online person re-identification services face privacy breaches from potential data leaks and recovery attacks, exposing cloud-stored images to malicious attackers and triggering public concern. The privacy protection of pedestrian images is crucial. Previous privacy-preserving person re-identification methods are unable to resist recovery attacks and compromise accuracy. In this paper, we propose an iterative method (PixelFade) to optimize pedestrian images into noise-like images to resist recovery attacks. We first give an in-depth study of protected images from previous privacy methods, which reveal that the \textbf{chaos} of protected images can disrupt the learning of recovery networks, leading to a decrease in the power of the recovery attacks. Accordingly, we propose Noise-guided Objective Function with the feature constraints of a specific authorization model, optimizing pedestrian images to normal-distributed noise images while preserving their original identity information as per the authorization model. To solve the above non-convex optimization problem, we propose a heuristic optimization algorithm that alternately performs the Constraint Operation and the Partial Replacement operation. This strategy not only safeguards that original pixels are replaced with noises to protect privacy, but also guides the images towards an improved optimization direction to effectively preserve discriminative features. Extensive experiments demonstrate that our PixelFade outperforms previous methods in resisting recovery attacks and Re-ID performance. The code will be released.

Primary Subject Area: [Experience] Multimedia Applications

Relevance To Conference: This work centers on the topic of Privacy-preserving Person Re-identification, a critical aspect of multimedia processing, which addresses the urgent need for privacy in handling pedestrian images. In the context of increasing surveillance and data collection, public concerns about privacy breaches in online Re-ID services are mounting, potentially hindering the development of Person Re-identification applications. To solve this, our goal is to safeguard the visual privacy of pedestrian images while maintaining their performance of re-identification models. We propose an iterative method, PixelFade, that innovatively transforms pedestrian images into noise-like images, effectively protecting visual information and resisting recovery attacks while preserving the original identity features for authorized models. Our extensive experiments demonstrate that our PixelFade outperforms existing privacy-preserving methods in terms of resisting recovery attacks and re-identification accuracy. We hope that our privacy-preserving pedestrian re-identification method can alleviate privacy concerns, thereby fostering the development of multimedia processing applications that benefit humanity.

Submission Number: 3149