Generative Unlearning for Any Identity

TL;DR

This paper introduces GUIDE, a novel method for generative identity unlearning that prevents models from generating images of specific identities while maintaining overall image quality, using minimal data.

Contribution

The paper presents GUIDE, a new framework for unlearning specific identities in generative models with only a single image, advancing privacy-preserving generative techniques.

Findings

Achieves state-of-the-art performance in identity unlearning.

Effectively prevents generation of specific identities.

Maintains high quality of generated images.

Abstract

Recent advances in generative models trained on large-scale datasets have made it possible to synthesize high-quality samples across various domains. Moreover, the emergence of strong inversion networks enables not only a reconstruction of real-world images but also the modification of attributes through various editing methods. However, in certain domains related to privacy issues, e.g., human faces, advanced generative models along with strong inversion methods can lead to potential misuses. In this paper, we propose an essential yet under-explored task called generative identity unlearning, which steers the model not to generate an image of a specific identity. In the generative identity unlearning, we target the following objectives: (i) preventing the generation of images with a certain identity, and (ii) preserving the overall quality of the generative model. To satisfy these goals, we propose a novel framework, Generative Unlearning for Any Identity (GUIDE), which prevents the reconstruction of a specific identity by unlearning the generator with only a single image. GUIDE consists of two parts: (i) finding a target point for optimization that un-identifies the source latent code and (ii) novel loss functions that facilitate the unlearning procedure while less affecting the learned distribution. Our extensive experiments demonstrate that our proposed method achieves state-of-the-art performance in the generative machine unlearning task. The code is available at https://github.com/KHU-AGI/GUIDE.