Reproduce “Sparse Global Matching for Video Frame Interpolation with Large Motion” 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
Chunxu Liu, Guozhen Zhang, Rui Zhao, Limin Wang
SGM-VFI fills missing video frames by first estimating motion locally, then using difference-map-guided points for global matching and merging the resulting compensation with the initial flows. On the challenging X-Test-L-4K subset, its Ours-1/2-Points model scores 29.91 PSNR versus 29.44 for a fine-tuned local-only version; this tests the added pipeline, not global matching alone, and BiFormer scores higher.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.