KAIST and Partners Unveil VPM, a Video World Model Grounded in 3D Point Trajectories

VPM from KAIST, ETH Zürich and others learns persistent 3D point trajectories to predict future multi-view video, acting as a 4D world model that improves both robotics performance and generation fidelity.

2026-10-21 ~ 2026-10-21 · 2 related posts

• Video-Point Model learns 4D world models via 3D point trajectory forecasting — Byungwoo Jeon · 2026-10-21
• VPM: teaching video world models to forecast 3D point trajectories improves videos and robots — sainingxie · 2026-10-21