PSyDUCK: Hiding Information in the Denoising Process of Latent Diffusion Models
Mahfuz, Aqib Channing, Georgia van der Wilk, Mark Torr, Philip HS Pizzati, Fabio de Witt, Christian Schroeder
Mahfuz, Aqib
Channing, Georgia
van der Wilk, Mark
Torr, Philip HS
Pizzati, Fabio
de Witt, Christian Schroeder
Supervisor
Department
Computer Vision
Embargo End Date
Type
Workshop
Date
License
Language
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Recent advances demonstrate that information can be covertly embedded in the outputs of stochastic generative AI models, raising both opportunities for secure communication and risks of misuse. Existing latent diffusion steganography methods typically hide data in the entropy of the initial latent state, inherently limiting embedding capacity. In this work, we instead investigate information hiding within the entropy of the diffusion denoising process itself. We introduce PSyDUCK, a simple but efficient framework that leverages controlled divergence and local mixing during denoising to enable high-capacity message embedding while preserving visual fidelity. Our empirical evaluation shows PSyDUCK can hide substantial information in both image and video diffusion models. While our formal analysis indicates that the security guarantees of denoising-based embedding are limited, the existence of this channel nonetheless requires that steganalysis methods account for entropy throughout the entire denoising process - not just in the initial latent state.
Citation
A. Mahfuz, G. Channing, M. van der Wilk, P.H.S. Torr, F. Pizzati, C.S. de Witt, "PSyDUCK: Hiding Information in the Denoising Process of Latent Diffusion Models," 2026, pp. 156-161.
Source
IEEE International Workshop on Information Forensics and Security Wifs
Conference
2025 IEEE International Workshop on Information Forensics and Security (WIFS)
Keywords
40 Engineering, 4006 Communications Engineering, 46 Information and Computing Sciences
Subjects
Source
2025 IEEE International Workshop on Information Forensics and Security (WIFS)
Publisher
IEEE