Hiding Images in Diffusion Models by Editing Learned Score Functions

Haoyu Chen, Yunqiao Yang, Nan Zhong, Kede Ma

Keywords: data hiding, steganography, diffusion models

Abstract:

Hiding data in deep neural networks (DNNs) has achieved remarkable successes, including both discriminative and generative models. Yet, the potential for hiding images in diffusion models remains underdeveloped. Existing approaches fall short in extracting fidelity, secrecy, and efficiency. In particular, the intensive computational demands of the hiding process, coupled with the slow extraction due to multiple denoising stages, make these methods impractical for resource-limited environments.
To address these challenges, we propose hiding images at a specific denoising stage in diffusion models by modifying the learned score functions. We also introduce a parameter-efficient fine-tuning (PEFT) approach that combines parameter selection with a variant of low-rank adaptation (LoRA) to boost secrecy and hiding efficiency.
Comprehensive experiments demonstrate the effectiveness of our proposed method.

Primary Area: other topics in machine learning (i.e., none of the above)

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 5658