IRIS embeds a keyed ring derived from the generated image's CLIP-based semantic code back into late diffusion steps, making the watermark fail automatically when the image's semantics change.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
IRIS: Visual-Semantic Binding for Forgery-Resistant Watermarking of Diffusion Images
IRIS embeds a keyed ring derived from the generated image's CLIP-based semantic code back into late diffusion steps, making the watermark fail automatically when the image's semantics change.