Towards Immersive Human-X Interaction: A Real-Time Framework for Physically Plausible Motion Synthesis
By: Kaiyang Ji , Ye Shi , Zichen Jin and more
Potential Business Impact:
Makes robots move and react like real people.
Real-time synthesis of physically plausible human interactions remains a critical challenge for immersive VR/AR systems and humanoid robotics. While existing methods demonstrate progress in kinematic motion generation, they often fail to address the fundamental tension between real-time responsiveness, physical feasibility, and safety requirements in dynamic human-machine interactions. We introduce Human-X, a novel framework designed to enable immersive and physically plausible human interactions across diverse entities, including human-avatar, human-humanoid, and human-robot systems. Unlike existing approaches that focus on post-hoc alignment or simplified physics, our method jointly predicts actions and reactions in real-time using an auto-regressive reaction diffusion planner, ensuring seamless synchronization and context-aware responses. To enhance physical realism and safety, we integrate an actor-aware motion tracking policy trained with reinforcement learning, which dynamically adapts to interaction partners' movements while avoiding artifacts like foot sliding and penetration. Extensive experiments on the Inter-X and InterHuman datasets demonstrate significant improvements in motion quality, interaction continuity, and physical plausibility over state-of-the-art methods. Our framework is validated in real-world applications, including virtual reality interface for human-robot interaction, showcasing its potential for advancing human-robot collaboration.
Similar Papers
Uni-Inter: Unifying 3D Human Motion Synthesis Across Diverse Interaction Contexts
CV and Pattern Recognition
Makes computer characters move realistically together.
PhysHMR: Learning Humanoid Control Policies from Vision for Physically Plausible Human Motion Reconstruction
CV and Pattern Recognition
Makes computer-animated people move like real humans.
A Whole-Body Motion Imitation Framework from Human Data for Full-Size Humanoid Robot
Robotics
Robots copy human moves, staying balanced.