Reproduce “UniSlider: Perceptually Uniform Sliders for Continuous Image Editing” 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
David Serrano-Lozano, Duygu Ceylan, Yannick Hold-Geoffroy, Iliyan Georgiev, Javier Vazquez-Corral, Anna Frühstück
UniSlider makes image-editing sliders more predictable by training an adapter to produce a smoother path between an input image and the editor’s default result, then remapping slider positions along that path. On a 300-example benchmark, it reaches a 0.652 coefficient of variation for consecutive-frame DreamSim distances and 0.799 VQA edit success; results do not guarantee human-perceived uniformity or correct edits.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.