Search papers, labs, and topics across Lattice.
ActiveGlasses introduces a system for learning robot manipulation from egocentric human demonstrations using smart glasses with a stereo camera as the sole perception device. Object trajectories are extracted from demonstrations to train an object-centric point-cloud policy that jointly predicts manipulation and head movement. The system achieves zero-shot transfer with active vision, outperforming baselines and generalizing across robot platforms in challenging tasks.
Robots can now learn complex manipulation tasks directly from human demonstrations using only a pair of smart glasses, achieving zero-shot transfer without specialized hardware.
Large-scale real-world robot data collection is a prerequisite for bringing robots into everyday deployment. However, existing pipelines often rely on specialized handheld devices to bridge the embodiment gap, which not only increases operator burden and limits scalability, but also makes it difficult to capture the naturally coordinated perception-manipulation behaviors of human daily interaction. This challenge calls for a more natural system that can faithfully capture human manipulation and perception behaviors while enabling zero-shot transfer to robotic platforms. We introduce ActiveGlasses, a system for learning robot manipulation from ego-centric human demonstrations with active vision. A stereo camera mounted on smart glasses serves as the sole perception device for both data collection and policy inference: the operator wears it during bare-hand demonstrations, and the same camera is mounted on a 6-DoF perception arm during deployment to reproduce human active vision. To enable zero-transfer, we extract object trajectories from demonstrations and use an object-centric point-cloud policy to jointly predict manipulation and head movement. Across several challenging tasks involving occlusion and precise interaction, ActiveGlasses achieves zero-shot transfer with active vision, consistently outperforms strong baselines under the same hardware setup, and generalizes across two robot platforms.
