Reproduce “Local Content-Style Control for Diffusion-based Image Stylization” 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
Amir Semmo
The authors make stylization locally controllable by turning ControlNet strength and IP-Adapter scale into spatial maps, allowing users to adjust structure preservation and style influence by region in a generation pass without retraining. In a 300-example image-stylization evaluation, relaxing style to 0.7 produced far-field change ratios of 0.26, 0.10, and 0.13; the controls predominantly steer separate axes, not perfectly independently.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.