Reproduce “PixelSpline: Continuous Space-Time Video Super-Resolution via Endpoint-Anchored Shared Pixel Trajectories” 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
Wanqi Yuan, Nianyi Li
PixelSpline performs continuous space-time video super-resolution using long-context features and a shared path to transport matched pixels between endpoint frames. On Adobe240 Average at 4× spatial and 8× temporal upsampling, it reports 31.39 dB PSNR and 406.41 ms per frame, versus V³ at 32.29 dB and 912.21 ms; these results are specific to that dataset and timing protocol.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.