Reproduce “LociArt: A Unified Paradigm for Free-Style Local Video 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

Yuxuan Chen, Feng Xie, Jiagao Hu, Fuhao Li, Zepeng Wang, Fei Wang, Daiguo Zhou

LociArt enables localized video stylization by encoding style identity in an edit mask and combining it with geometric cues and source-video evidence. It also introduces training data and a benchmark evaluating style, preservation, and temporal consistency separately. On 168 benchmark cases, GRNStyle-2B reaches 0.705 background SPS versus 0.676 for fine-tuned VACE; this does not establish superiority or boundary stability.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.