Reproduce “Gaussian Image Steganography via Parameter-Domain Keyed Embeddings” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tong Wu, Runze Cheng, Xiaoyue Fan, Kaan Akşit
The method hides an 8-bit message in selected parameters of a fitted image-space 2D Gaussian representation. A secret key guides cost-aware edits to shape, opacity, and luminance while preserving the render. On 112 held-out natural images, the correct key recovered every payload, with wrong-key mean BER 0.520; decoding requires paired parameters, clean references, and an assignment map.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.