PRIME: Protect Your Videos From Malicious Editing

• DOI: 10.48550/arxiv.2402.01239
• Publication Date: 2024-02-02
• Authors: Guanlin Li; Shuai Yang; Jie Zhang; Tianwei Zhang

With the development of generative models, quality generated content keeps increasing. Recently, open-source models have made it surprisingly easy to manipulate and edit photos videos, with just a few simple prompts. While these cutting-edge technologies gained popularity, they also given rise concerns regarding privacy portrait rights individuals. Malicious users can exploit tools for deceptive or illegal purposes. Although some previous works focus on protecting against we find there are still gaps between videos images in aspects efficiency effectiveness. Therefore, introduce our protection method, PRIME, significantly reduce time cost improve performance. Moreover, evaluate proposed consider both objective metrics human subjective metrics. Our evaluation results indicate that PRIME only costs 8.3% GPU hours state-of-the-art method achieves better Code be found https://github.com/GuanlinLee/prime.