The authors recorded egocentric human demonstrations and translated the observed full-body motions into actions executable by a humanoid.
Human demonstrations teach humanoids coordinated whole-body manipulation without teleoperation
EgoHumanoid-V2 aligns egocentric human motion with humanoid dynamics, enabling vision-language-action policies to perform four real-world tasks without task-specific robot demonstrations.
Chinese Tech
Jin Chen · Yiming Jiang · Chongyang Xu · Modi Shi · Shijia Peng · Li Chen · +5 more
OpenDriveLab at The University of Hong Kong · Alibaba Group · Shanghai Innovation Institute · Fudan University · Beihang University
Research Digest··2 min read
Chen and colleagues present a pipeline for converting egocentric human demonstrations into training data for coordinated humanoid loco-manipulation, where locomotion, balance and manipulation must work together.
Why this paper
From Alibaba Group and 6 others
In one line
EgoHumanoid-V2 transfers coordinated whole-body loco-manipulation skills from egocentric human demonstrations to humanoid robots.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors (2 noted)
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§