Search papers, labs, and topics across Lattice.
OmniClone, a whole-body humanoid teleoperation system, was developed and evaluated using a new diagnostic benchmark, OmniBench, which assesses performance across various motion categories and difficulty levels. The system incorporates subject-agnostic retargeting and robust communication to improve performance and generalization. Results show a 66% reduction in Mean Per-Joint Position Error (MPJPE) with significantly fewer computational resources, and the system supports real-time teleoperation, motion playback, and VLA model control.
A new humanoid teleoperation system, OmniClone, achieves state-of-the-art control fidelity and generalization while drastically reducing computational cost, thanks to a diagnostic benchmark that exposes failure modes overlooked by aggregate metrics.
Whole-body humanoid teleoperation enables humans to remotely control humanoid robots, serving as both a real-time operational tool and a scalable engine for collecting demonstrations for autonomous learning. Despite recent advances, existing systems are validated using aggregate metrics that conflate distinct motion regimes, masking critical failure modes. This lack of diagnostic granularity, compounded by tightly coupled and labor-intensive system configurations, hinders robust real-world deployment. A key open challenge is building a teleoperation system that is simultaneously robust, versatile, and affordable for practical use. Here we present OmniClone, a whole-body humanoid teleoperation system that achieves high-fidelity, multi-skill control on a single consumer GPU with modest data requirements. Central to our approach is OmniBench, a diagnostic benchmark that evaluates policies across stratified motion categories and difficulty levels on unseen motions, exposing the narrow specialization of prior systems. Guided by these diagnostics, we identify an optimized training data recipe and integrate system-level improvements: subject-agnostic retargeting and robust communication, that collectively reduce Mean Per-Joint Position Error (MPJPE) by over 66% while requiring orders-of-magnitude fewer computational resources than comparable methods. Crucially, OmniClone is control-source-agnostic: a single unified policy supports real-time teleoperation, generated motion playback, and Vision-Language-Action (VLA) models, while generalizing across operators of vastly different body proportions. By uniting diagnostic evaluation with practical engineering, OmniClone provides an accessible foundation for scalable humanoid teleoperation and autonomous learning.
