DyRAD: radar novel view synthesis renders full range-azimuth-Doppler tensors for dynamic driving scenes
orlitany · x · 2026-10-02
Researchers from Technion, Cornell Tech, and NVIDIA present DyRAD, a radar novel view synthesis method for dynamic driving scenes, with paper, project page, and code.
- Radar senses motion directly via Doppler, but prior methods either ignore dynamics or assume static scenes.
- DyRAD models scenes with static background reflectors plus motion-tracked dynamic point reflectors, rendering full range-azimuth-Doppler (RAD) tensors; Doppler serves as both output and supervision.
- A fixed analytic point-spread function derived from the radar's signal-processing chain prevents sensor-induced spread from baking into scene geometry.
- Enables novel viewpoint rendering, scene editing, and sensor reconfiguration for closed-loop autonomous driving evaluation.
Related event: DyRAD Enables Radar Novel View Synthesis for Dynamic Driving(4 posts)→
More from Research
- AI 'speech clock' predicts how fast you're ageing from your voice, featured in Nature — AnnaCiaunica · 2026-10-02
- ScholarCatalyst: a new benchmark testing whether AI can pick research problems like humans — PangWeiKoh · 2026-10-02
- Clarification as Supervision lands NeurIPS Oral: denser training signal via model interaction — iatitov · 2026-10-02
- Ego-Exo4D-HM: SMPL-H reconstructions for 523 hours of egocentric-exocentric video, open-sourced — geopavlakos · 2026-10-02
- Cohere Labs Talk: Making AI Math Reasoning Machine-Checkable with Lean — Cohere_Labs · 2026-10-02
- New paper dissects only task-relevant network parts, making mechanisms inspectable and editable at far lower cost — leedsharkey · 2026-10-02