WMAdapter: Adding WaterMark Control to Latent Diffusion Models

Hai Ci, Yiren Song, Pei Yang, Jinheng Xie, Mike Zheng Shou

TL;DR: Watermark LDM via a plugin without compromising image quality

Abstract: Watermarking is essential for protecting the copyright of AI-generated images. We propose WMAdapter, a diffusion model watermark plugin that embeds user-specified watermark information seamlessly during the diffusion generation process. Unlike previous methods that modify diffusion modules to incorporate watermarks, WMAdapter is designed to keep all diffusion components intact, resulting in sharp, artifact-free images. To achieve this, we introduce two key innovations: (1) We develop a contextual adapter that conditions on the content of the cover image to generate adaptive watermark embeddings. (2) We implement an additional finetuning step and a hybrid finetuning strategy that suppresses noticeable artifacts while preserving the integrity of the diffusion components. Empirical results show that WMAdapter provides strong flexibility, superior image quality, and competitive watermark robustness.

Lay Summary: AI-generated images are increasingly at risk of misuse, raising concerns about copyright protection. We created WMAdapter, a tool that invisibly marks images during generation without affecting their quality. This helps creators prove ownership and protect their work, offering a simple and effective solution for safeguarding AI-generated content.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Link To Code: https://github.com/showlab/WMAdapter

Primary Area: Deep Learning->Generative Models and Autoencoders

Keywords: watermark, latent diffusion model

Submission Number: 1914