URDF Synthesis from RGB-D Sequences via Differentiable Joint Inference and Energy-Consistent Verification
arXiv:2606.18861v1 Announce Type: cross Abstract: Reconstructing simulation-ready digital twins of articulated objects from sensor observations remains constrained by two persistent gaps: (i) part-level geometric reconstruction is decoupled from kinematic-parameter estimation, and (ii) the recovered models often violate basic dynamic invariants such as energy conservation, leading to drift when the URDF is replayed in physics simulators. We present KinemaForge, a constraint-driven pipeline that jointly infers part-level shape, joint topology, and joint parameters from short RGB-D sequences and
The proliferation of RGB-D sensors and advancements in differentiable rendering and joint inference are enabling more sophisticated 3D reconstruction techniques, making this breakthrough timely.
This development addresses critical limitations in creating accurate digital twins for articulated objects, which is foundational for progress in robotics, simulation, and virtual environments.
The ability to jointly infer shape, joint topology, and parameters from sensor data, coupled with energy-consistent verification, significantly improves the fidelity and usability of URDF models in physics simulators.
- · Robotics companies
- · Simulation software developers
- · Digital twin providers
- · AI/ML researchers in computer vision
- · Companies relying on manual 3D model creation
- · Less accurate simulation tools
More robust and accurate robotic manipulation and assembly simulations will become possible.
Reduced development time and cost for new robotic systems, leading to faster innovation cycles.
Enhanced automation in manufacturing and logistics through more reliable digital representations of objects and environments.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI