Peking University Open-Sources PanDWM, a 360° 4D Driving World Model Beating DiST-4D
PekingUniversity · hf · 2026-11-18
Peking University released PanDWM (Panoramic 4D Driving World Models) on Hugging Face:
- Stitches six camera views into a 360° panoramic band; a distilled causal DiT (SD3-based, 4-step) generates panoramic driving video conditioned on panoramic history, 3D boxes + HD map, text, and virtual-camera ray maps;
- A feed-forward 3D Gaussian decoder lifts latents into pixel-aligned Gaussians, enabling novel-view rendering and 4D static completion (e.g., move 5m forward, turn 30° left);
- Outperforms multi-camera baselines like DiST-4D with faster generation and better FID/FVD, with notably more consistent long-horizon rollouts.
Useful for autonomous driving simulation and world-model research.
More from Embodied
- RoboDojo sim "not even wrong": blog exposes physics flaws in popular robot benchmark — ZeYanjie · 2026-11-18
- NVIDIA-backed Agile ONE SN humanoid robot for researchers ships early 2027 — xmercury_one · 2026-11-18
- Humanoids serve ice cream and spar in Shenzhen while APEC leaders visit — TansuYegen · 2026-11-18
- Otto by Rockwell explains using Gazebo digital twins for AMR deployments — ___Mufasaa · 2026-11-18
- Tokkey debuts thumb-sized AI hardware to run local models on your Mac — TejasKumar_ · 2026-11-18
- 'Nothing more physical AI than moving tons of rock': live demo at Web Summit shows real-world difficulty — MarvinTBaumann · 2026-11-18