Transferring the Intelligence of VLMs to Robotic Control
arXiv preprint, 2026
Studies whether the intelligence of vision-language models can carry over from the digital world to physical robotic control.
TopoCap: Learning Topology-Agnostic Motion Priors for Monocular Video-to-Animation
SIGGRAPH 2026 Conference Papers
Captures motion from monocular video and retargets it onto characters with arbitrary, unseen skeletal topologies, from bipeds to hexapods, without test-time optimization.
SkinTokens: A Learned Compact Representation for Unified Autoregressive Rigging
arXiv preprint, 2026
A discrete, compact representation of skinning weights that turns rigging into token prediction, unifying skeleton and skinning generation in one autoregressive model.
One Model to Rig Them All: Diverse Skeleton Rigging with UniRig
ACM Transactions on Graphics (SIGGRAPH 2025)
A single model that automatically rigs diverse 3D assets, from humans and animals to inanimate objects, with skeletons and skinning weights.
CharacterGen: Efficient 3D Character Generation from Single Images with Multi-View Pose Canonicalization
ACM Transactions on Graphics (SIGGRAPH 2024)
Generates a pose-canonicalized 3D character mesh from a single image, ready for downstream rigging and animation.