RoSHI: A Versatile Robot-oriented Suit for Human Data In-the-Wild

Wenjing Margaret Mao, Jefferson Ng, Luyang Hu, Daniel Gehrig, Antonio Loquercio · 2026

Read RoSHI: A Versatile Robot-oriented Suit for Human Data In-the-Wild by Wenjing Margaret Mao, Jefferson Ng, Luyang Hu, Daniel Gehrig, Antonio Loquercio (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-07331.txt.

Description

Mao and four colleagues introduce RoSHI, a wearable suit for capturing human motion in the wild to train robots. It fuses cheap sparse motion sensors with Project Aria glasses, combining occlusion-proof body tracking with egocentric mapping that anchors long motions in a global frame. The system estimates full 3D pose and body shape from the wearer's own viewpoint. On agile activities it beats other egocentric baselines, matches a leading external-camera baseline, and yields motion data usable for real humanoid policies.

Catalog record

CREATORWenjing Margaret Mao, Jefferson Ng, Luyang Hu, Daniel Gehrig, Antonio Loquercio
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE31131
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.07331

Browse the full library · About the Tower of Babel