Reproduce “Neural Video Coding under Limited Delay: Theoretical Optimality and System Design” 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
Likun Cai, Jun Chen, Ashish Khisti
GFSC lets encoders use a window of future frames when coding a frame. Its theory shows that a K-frame delay reaches the joint-coding rate–distortion function for finite-alphabet, positive-distribution K-order Markov sources over a nonempty region of small positive distortions. In experiments, DCVC-GFSC with K=1 achieves −20.57% BD-rate on UVG 1080p versus zero-delay DCVC; this result is limited to that comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.