Hesheng Wang

PhyVisGen, a physically and visually high-fidelity framework for scalable robotic manipulation data generation, introduces an arm-gripper coupling method based on the Incremental Potential Contact (IPC), enabling high-fidelity soft contact throughout complete manipulation trajectories.

DynaForge combines low-frequency global planning with high-frequency object-centric inverse kinematics across task phases, and applies a residual policy to correct actions during dynamic interaction, showing the ability of DynaForge for sim-to-real transfer.

This work investigates compositional reuse at two levels: within a subtask, combining familiar entities, action types, and action modifiers; and across subtasks, reusing learned subtasks in unseen long-horizon tasks, which validate the effectiveness of the method for instruction-dependent generalization at both levels.

Achieving state-of-the-art robust estimation efficiency, DiffSAC slashes hypothesis evaluations from over 10,000 to just dozens.

Robots can now learn new tasks on-the-fly without retraining, drastically cutting down the time and resources needed for adaptation.

Achieving state-of-the-art performance in scene understanding and controllable 4D generation, GaussianDWM++ redefines how we integrate language and 3D scene editing.