Attention to Neural Plagiarism: Diffusion Models Can Plagiarize Your Copyrighted Images!

TL;DR

A general approach to neural plagiarism that can either forge replicas of copyrighted data or introduce copyright ambiguity is proposed, based on anchors and shims, which employs inverse latents as anchors and finds shim perturbations that gradually deviate the anchor latents, thereby evading watermark or copyright detection.

Abstract

In this paper, we highlight a critical threat posed by emerging neural models: data plagiarism. We demonstrate how modern neural models (e.g., diffusion models) can replicate copyrighted images, even when protected by advanced watermarking techniques. To expose vulnerabilities in copyright protection and facilitate future research, we propose a general approach to neural plagiarism that can either forge replicas of copyrighted data or introduce copyright ambiguity. Our method, based on "anchors and shims", employs inverse latents as anchors and finds shim perturbations that gradually deviate the anchor latents, thereby evading watermark or copyright detection. By applying perturbations to the cross-attention mechanism at different timesteps, our approach induces varying degrees of semantic modification in copyrighted images, enabling it to bypass protections ranging from visible trademarks and signatures to invisible watermarks. Notably, our method is a purely gradient-based search that requires no additional training or fine-tuning. Experiments on MS-COCO and real-world copyrighted images show that diffusion models can replicate copyrighted images, underscoring the urgent need for countermeasures against neural plagiarism.

Figures (2)

• Figure 1: Copyrighted images (top left) and watermarked data (bottom left) can be easily plagiarized by diffusion models.
• Figure 2: Attack pipeline overview for Neural plagiarism.