SOURCE-LINKED INTELLIGENCE

Training-Free Inpainting Across Domains with a Frozen Text-to-Image Diffusion Model

arXiv · AI, language, vision and robotics · article · Sep 1, 2026 · UTC

We show that a frozen generic text-to-image diffusion model can perform conditional inpainting across three evaluated natural-image domains with one fixed controller configuration, without inpainting-specific weight training, dataset-specific weight adaptation, or learned inpainting-specific conditioning channels. Step-PI augments known-region projection with boundary-interior latent feedback, persistent PI state, and a predefined four-field release schedule that modulates controller signals along the reverse trajectory. Developed only on Main35-disjoint CelebA-HQ pilots, the controller transf

Read original source ↗ Open in workspace

recordType

paper

region

Global