MIT's Point2Pose tracks unknown objects in 6D with full-occlusion recovery
joemeno · x · 2026-09-10
MIT's Point2Pose (Tzu-Yuan Lin et al., ECCV 2026) performs CAD-free 6D pose tracking and 3D reconstruction of multiple unknown objects. Using 2D point tracking, it instantly recovers poses even after full occlusion, with no object model required. The team also released the YCBMultiTrack evaluation dataset.
More from Embodied
- Freeform Preference Learning Boosts Robot Manipulation Over Baselines by 38 Percentage Points — StanfordAILab · 2026-09-10
- BHF Robotics' Gen3 kills weeds by electrocuting roots, no herbicide needed — chrisgrayson · 2026-09-10
- EPIC-Contact dataset and HOPformer bring 3D hand-object pose estimation into the wild — Michael_J_Black · 2026-09-10
- Open-source tabletop DIY home AI agent built on openclaw and Raspberry Pi — steipete · 2026-09-10
- SyncWorld turns world models into zero-shot robot simulators via visual calibration — Yuncong Yang · 2026-09-10
- Reservoir Farms to showcase crop robots in Salinas on Sept 25 — Scobleizer · 2026-09-10