OmniInteract improves dual humanoid whole-body interaction control
OmniInteract: Towards Unified Whole-Body Interaction Control for Dual Humanoids
Robotics
Summary
Working together physically like robots do is hard because their shapes and movements vary a lot. The authors created OmniInteract, which helps two robot bodies mimic human contact and coordinate better by using detailed surface matching and smarter motion adjustments. It keeps natural contact points and adapts motions to each partner, even as the types of interactions grow. Their method shows better teamwork between two robots and recovers well if pushed or disturbed.
What this means in practice
- •For robotics engineers: Enable dual humanoid robots to physically collaborate with stable whole-body contact and coordinated motion in complex tasks.
- •For robotics middleware developers: Create control modules that unify diverse dual-robot interaction skills while preserving underlying motor capabilities.
Authors
Yinghao Li, Tianyu Zhang, Zhuohao Li, Yi-Lin Wei, Fa-Ting Hong, Qing Zhang, Wei-Shi Zheng
Abstract
Physical interaction is a key capability for humanoid collaboration, requiring coordinated whole-body motion and precise contact. However, learning diverse dual-humanoid interactions faces two major challenges: (1) human-robot morphological differences disrupt contact interfaces and spatial relationships during retargeting, and (2) jointly learning motor skills and partner coordination makes it difficult to retain control fidelity as interaction diversity grows. In this paper, we propose OmniInteract to address these challenges through surface-aware modeling and scalable residual policy learning. First, we introduce Surface Contact-Aware Retargeting, which leverages dense human-robot surface correspondences to jointly retarget both humanoids, preserving explicit interaction contact interfaces and their near-field geometric context. Second, we propose a Scalable Flow-based Interaction Policy for partner-conditioned residual generation. It factorizes dual-humanoid control into a reusable motor foundation and interaction-specific adaptations: partner-aware experts learn corrections over a frozen whole-body motion prior, while conditional flow matching unifies them in a shared residual action space. This enables scalable skill consolidation while retaining pretrained motor capabilities. Extensive experiments demonstrate that OmniInteract improves retargeting fidelity and dual-humanoid coordination while better retaining interaction performance as skill diversity grows. Real-world deployment further demonstrates successful contact establishment, stable whole-body coordination, and recovery from external perturbations.