Jiange Yang

Tra-MoE: learning trajectory prediction model from multiple domains for adaptive policy conditioning

SPA: 3D Spatial-Awareness Enables Effective Embodied Representation