Reproduce “Sekai: Incremental Event Memory for Generalized Reasoning Video Segmentation” 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
Zhengtao Liu, Jinrong Xu, Shiu-Hong Kao, Huixin Sun, Yicong Li
Sekai addresses online video segmentation queries whose target set changes as events occur. It records query-relevant object events, then uses that record to select currently visible objects and segment them. On Isekai, it scores 50.33 J&F, versus 47.16 for adapted CoT-RVS. Its clearest advantage is on EVER, while stale records hurt LATEST and NEVER.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.