RigidFormer accepted at NeurIPS: mesh-free Transformer beats GNN simulators for rigid dynamics
camenduru · x · 2026-10-14
- RigidFormer (Learning Rigid Dynamics using Transformers) is accepted at NeurIPS 2026, with code open-sourced.
- It's an object-centric, mesh-free Transformer that learns multi-object rigid-body contact dynamics from point clouds, without traditional physics engines, matching or beating mesh-based baselines.
- The team (incl. MIT's Wojciech Matusik) reports higher accuracy and faster simulation than prior GNN-based SOTA, and hopes the rigid-alignment component extends to deformable-body dynamics.
More from Embodied
- Dev ports Wright CAD workflow to ChatGPT Desktop, wiring Codex to SteveCAD for 3D design — burhop · 2026-10-14
- RoboStrategy claims the first AI robot to ring the Nasdaq bell, $BOT draws attention — cixliv · 2026-10-14
- Where Winds Meet to add NVIDIA DLSS 5 with 3D-Guided Neural Rendering in late November — ctnzr · 2026-10-14
- LightNav-0 open-sourced: one VLM-based model navigates four robot types autonomously — rsasaki0109 · 2026-10-14
- Science Corporation implants vision-restoring chip in first Stargardt patient — TechCrunch AI · 2026-10-14
- Mind Robotics, Rivian's robotics spinoff with $1B raised, hires ex-Ford EV chief Doug Field — Sethwinterroth · 2026-10-14