Reproduce “DyRAD: Radar Novel View Synthesis for Dynamic Driving Scenes” 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
Merav Keidar, Tomer Borreda, Rajalakshmi Nandakumar, Or Litany
DyRAD models radar scenes for rendering recordings from new viewpoints. It separates fixed background reflectors from tracked, annotated moving objects, projects relative motion onto Doppler, and applies a fixed, sensor-specific measurement response. On held-out, on-path RADIal frames, its foreground hit rate is 0.907 ± 0.086, versus 0.269 ± 0.245 for RadarFields; this measures reference-occupied boxes, not accuracy at unseen positions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.