EgoHumanoid-V2: Human-to-Humanoid Transfer of Coordinated Whole-Body Skills for Loco-Manipulation

Jin Chen, Yiming Jiang, Chongyang Xu, Modi Shi, Shijia Peng, Li Chen, Tianyu Li, Mu Xu, Yilun Chen, Steven Hoi, Hongyang Li

arXiv preprint, 2026

Project | arXiv

_{Transfers coordinated whole-body skills from egocentric human demonstrations through coarse-to-fine action alignment and visual alignment, with zero-shot deployment on four real-world tasks.}

Towards a General Humanoid Loco-Manipulation Model via Egocentric Whole-Body Human Data Pretraining

Chongyang Xu, Zhao Wu, Jin Chen, Yiming Jiang, Jinhui Ye, Yuming Jiang, Shifeng Zhang, Ziliang Feng, Mu Xu, Yilun Chen, Li Lu, Steven C. H. Hoi

arXiv preprint, 2026

Project | GitHub | arXiv

_{Introduces HumanVerse-500, a 500-hour egocentric whole-body dataset, and lambda-zero, a humanoid VLA policy that transfers human experience through three-stage training.}

EgoAlign: Bridging the Human-Humanoid Gap for Long-Range Loco-Manipulation

Yiming Jiang, Jin Chen, Chongyang Xu, Yilun Chen, Aimin Hao, Yisheng He

arXiv preprint, 2026

Project | GitHub | arXiv

_{Aligns human demonstrations with humanoid actions and states, enabling zero-shot long-range loco-manipulation without physical-robot task demonstrations.}

EgoHumanoid: Unlocking In-the-Wild Loco-Manipulation with Robot-Free Egocentric Demonstration

Modi Shi^*, Shijia Peng^*, Jin Chen^*, Haoran Jiang, Yinghui Li, Di Huang, Ping Luo, Hongyang Li, Li Chen

Robotics: Science and Systems (RSS), 2026

Project | GitHub | arXiv

_{The first framework that trains on robot-free egocentric human videos and zero-shot deploys on bipedal humanoid robots.}

χ₀: Resource-Aware Robust Manipulation via Taming Distributional Inconsistencies

Checheng Yu, Chonghao Sima, Gangcheng Jiang, Hai Zhang, Haoguang Mai, Hongyang Li, Huijie Wang, Jin Chen, Kaiyang Wu, Li Chen, Lirui Zhao, Modi Shi, Ping Luo, Qingwen Bu, Shijia Peng, Tianyu Li, Yibo Yuan   _{(authors listed alphabetically)}

RSS 2026 Workshop on Beyond Rigidity, 2026

Project | GitHub | arXiv

_{A systematic study and optimization of dual-arm dexterous deformable-object manipulation; powered the first 24-hour non-stop laundry-folding livestream.}

WholeBodyVLA: Towards Unified Latent VLA for Whole-Body Loco-Manipulation Control

Haoran Jiang^*, Jin Chen^*, Qingwen Bu, Li Chen, Modi Shi, Yanjie Zhang, Delong Li, Chuanzhe Suo, Chuang Wang, Zhihui Peng, Hongyang Li

International Conference on Learning Representations (ICLR), 2026

Project | GitHub | arXiv

_{The first end-to-end VLA autonomy stack for bipedal humanoids, and the first to sustain over 1 minute of autonomous loco-manipulation on a bipedal humanoid.}

Is Diversity All You Need for Scalable Robotic Manipulation?

Modi Shi^*, Li Chen^*, Jin Chen^*, Yuxiang Lu, Chiming Liu, Guanghui Ren, Ping Luo, Di Huang, Maoqing Yao, Hongyang Li

IEEE Transactions on Robotics (T-RO), 2026

Project | GitHub | arXiv

_{A scaling-law study of how VLA foundation model performance scales with real-robot data diversity.}

The 1st AgiBot World Challenge

IROS 2025 Competition, Tech Lead

Two tracks — Manipulation and World Model — attracting 300+ teams worldwide. Built on the AgiBot World platform with real-robot evaluation pipelines.

Website