Latent Diffusion Models for Attribute-Preserving Image Anonymization

arXiv:2403.14790v1 Announce Type: new
Abstract: Generative techniques for image anonymization have great potential to generate datasets that protect the privacy of those depicted in the images, while achieving high data fidelity and utility. Existing methods have focused extensively on preserving facial attributes, but failed to embrace a more comprehensive perspective that considers the scene and background into the anonymization process. This paper presents, to the best of our knowledge, the first approach to image anonymization based on Latent Diffusion Models …

abstract anonymization arxiv cs.ai cs.cv data datasets diffusion diffusion models fidelity generate generative image images latent diffusion models perspective privacy protect type utility