GeoState: Learning View-Decoupled Geometry States for 3D World Modeling
GeoState decouples scene geometry from camera motion for predictive 3D world modeling.
3 stories, newest first · one card per story
GeoState decouples scene geometry from camera motion for predictive 3D world modeling.
Drive3R streams block-wise 3D reconstruction for driving and sets new nuScenes, KITTI, and Waymo results.
Query-based cost learning over reachable ego futures improves end-to-end driving collision rates on nuScenes and real logs.
Today 30 readers · 30 views · 1 read · all time 108K views