Hey That's Mine Imperceptible Watermarks are Preserved in Diffusion Generated Outputs

TL;DR

This paper demonstrates that imperceptible watermarks embedded in training data are preserved in images generated by diffusion models, enabling detection and protection of content ownership.

Contribution

It introduces a method to embed imperceptible watermarks in training data and shows these watermarks persist in generated images, facilitating content ownership verification.

Findings

Watermarks are preserved in generated images.

Watermarks correlate with specific features of training data.

Statistical tests can identify models trained on watermarked data.

Abstract

Generative models have seen an explosion in popularity with the release of huge generative Diffusion models like Midjourney and Stable Diffusion to the public. Because of this new ease of access, questions surrounding the automated collection of data and issues regarding content ownership have started to build. In this paper we present new work which aims to provide ways of protecting content when shared to the public. We show that a generative Diffusion model trained on data that has been imperceptibly watermarked will generate new images with these watermarks present. We further show that if a given watermark is correlated with a certain feature of the training data, the generated images will also have this correlation. Using statistical tests we show that we are able to determine whether a model has been trained on marked data, and what data was marked. As a result our system offers a solution to protect intellectual property when sharing content online.