Once-for-all: Efficient Visual Face Privacy Protection via Person-specific Veils

Zixuan Yang, Yushu Zhang, Tao Wang, Zhongyun Hua, Zhihua Xia, Jian Weng

Abstract: As billions of face images stored on cloud platforms contain sensitive information to human vision, the public confronts substantial threats to visual face privacy. In response, the community has proposed some perturbation-based schemes to mitigate visual privacy leakage. However, these schemes need to generate a new protective perturbation for each image, failing to satisfy the real-time requirement of cloud platforms. To address this issue, we present an efficient visual face privacy protection scheme by utilizing person-specific veils, which can be conveniently applied to all images of the same user without regeneration. The protected images exhibit significant visual differences from the originals but remain identifiable to face recognition models. Furthermore, the protected images can be recovered to originals under certain circumstances. In the process of generating the veils, we propose a feature alignment loss to promote consistency between the recognition outputs of protected and original images with approximate construction of feature subspace. Meanwhile, the block variance loss is designed to enhance the concealment of visual identity information. Extensive experimental results demonstrate that our scheme can significantly eliminate the visual appearance of original images and almost has no impact on face recognition models.

Primary Subject Area: [Systems] Systems and Middleware

Secondary Subject Area: [Experience] Multimedia Applications

Relevance To Conference:

The multimedia data, particularly face images contain a plethora of sensitive visual information to human vision, which poses a substantial visual face privacy threat to the public. In this paper, we propose an efficient visual face privacy protection by utilizing person-specific veils to meet the requirements of the real-time systems. The protected images are identifiable to a face recognition model but unrecognized to human vision.

The main contributions are as follows:

1. Our scheme enables a user to apply a person-specific veil to all his/her images to conveniently hide visual identity, without crafting new veils multiple times.

2. Our proposed method fosters the consistency between the recognition outputs of protected and original images by employing feature subspace and enhances the concealment of visual identity information with block variance loss.

Submission Number: 3624