Shuran Song

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

Geometry-aware 4D Video Generation for Robot Manipulation

DexMachina: Functional Retargeting for Bimanual Dexterous Manipulation

Steerable Video Action Model

Latent Policy Barrier: Learning Robust Visuomotor Policies by Staying In-Distribution

Rectified Point Flow: Generic Point Cloud Pose Estimation

DynaGuide: Steering Diffusion Polices with Active Dynamic Guidance

Compliant Residual DAgger: Improving Real-World Contact-Rich Manipulation with Human Corrections

Efficient Part-level 3D Object Generation via Dual Volume Packing

Language models scale reliably with over-training and on downstream tasks

A Practical Guide for Incorporating Symmetry in Diffusion Policy

Real2Code: Reconstruct Articulated Objects via Code Generation

Should VLMs be Pre-trained with Image Data?