GaussMarker embeds watermarks in both the spatial and frequency domains of initial diffusion noise and adds a learned restorer, reporting near-perfect detection across eight distortions and four attacks on three Stable Diffusion versions.
Compared to Stable Signature, LaWa is more robust to image modifications and can handle mul- tiple users for one image generation service
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
GaussMarker: Robust Dual-Domain Watermark for Diffusion Models
GaussMarker embeds watermarks in both the spatial and frequency domains of initial diffusion noise and adds a learned restorer, reporting near-perfect detection across eight distortions and four attacks on three Stable Diffusion versions.