WMAdapter: Adding WaterMark Control to Latent Diffusion Models

Hai Ci, Yiren Song, Pei Yang, Jinheng Xie, Mike Zheng Shou
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:10901-10919, 2025.

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-ci25a, title = {{WMA}dapter: Adding {W}ater{M}ark Control to Latent Diffusion Models}, author = {Ci, Hai and Song, Yiren and Yang, Pei and Xie, Jinheng and Shou, Mike Zheng}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {10901--10919}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/ci25a/ci25a.pdf}, url = {https://proceedings.mlr.press/v267/ci25a.html}, 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.} }
Endnote
%0 Conference Paper %T WMAdapter: Adding WaterMark Control to Latent Diffusion Models %A Hai Ci %A Yiren Song %A Pei Yang %A Jinheng Xie %A Mike Zheng Shou %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-ci25a %I PMLR %P 10901--10919 %U https://proceedings.mlr.press/v267/ci25a.html %V 267 %X 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.
APA
Ci, H., Song, Y., Yang, P., Xie, J. & Shou, M.Z.. (2025). WMAdapter: Adding WaterMark Control to Latent Diffusion Models. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:10901-10919 Available from https://proceedings.mlr.press/v267/ci25a.html.

Related Material