PulseAugur
EN
LIVE 17:45:19

New pipeline enables humanoid robots to manipulate objects while walking

By PulseAugur Editorial · [3 sources] ·

Researchers have developed CoorDex, a novel learning pipeline that enables humanoid robots to perform dexterous manipulation while in motion. This system converts high-dimensional body and hand control into coordinated latent residual control, allowing robots like the Unitree G1 to grasp bottles, open fridge doors, and manipulate cubes without stopping. The approach utilizes simulated demonstrations to train motion tracking teachers, distills them into latent priors, and then uses these priors for reinforcement learning, outperforming simpler methods in complex loco-manipulation tasks. AI

IMPACT Advances in humanoid robot manipulation capabilities, enabling more complex real-world tasks and interactions.

RANK_REASON The cluster consists of multiple research papers detailing new methods and datasets for humanoid robot manipulation.

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

New pipeline enables humanoid robots to manipulate objects while walking

COVERAGE [3]

  1. arXiv cs.AI TIER_1 English(EN) · Mingyu Ding ·

    CoorDex: Coordinating Body and Hand Priors for Continuous Dexterous Humanoid Loco-Manipulation

    Humanoid loco-manipulation is often simplified into a stop-and-go process: walking to an object, stopping to manipulate it, and then resuming locomotion. It also commonly relies on low degree-of-freedom (DoF) end effectors that behave like an open-close grasp primitive. We introd…

  2. arXiv cs.LG TIER_1 English(EN) · Zhenyu Zhao, Hongyi Jing, Xiawei Liu, Jiageng Mao, Abha Jha, Hanwen Yang, Rong Xue, Sergey Zakharov, Vitor Guizilini, Yue Wang ·

    Humanoid Everyday: A Comprehensive Robotic Dataset for Open-World Humanoid Manipulation

    arXiv:2510.08807v2 Announce Type: replace-cross Abstract: From loco-motion to dextrous manipulation, humanoid robots have made remarkable strides in demonstrating complex full-body capabilities. However, the majority of current robot learning datasets and benchmarks mainly focus …

  3. arXiv cs.AI TIER_1 English(EN) · Lujie Yang, Xiaoyu Huang, Zhen Wu, Angjoo Kanazawa, Pieter Abbeel, Carmelo Sferrazza, C. Karen Liu, Rocky Duan, Guanya Shi ·

    OmniRetarget: Interaction-Preserving Data Generation for Humanoid Whole-Body Loco-Manipulation and Scene Interaction

    arXiv:2509.26633v3 Announce Type: replace-cross Abstract: A dominant paradigm for teaching humanoid robots complex skills is to retarget human motions as kinematic references to train reinforcement learning (RL) policies. However, existing retargeting pipelines often struggle wit…

RELATED ENTITIES

RELATED TOPICS