[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81737-en":3,"doc-seo-81737-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},81737,4810365810221,"Aurora","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Learning Dexterous Manipulation Using Contact Wrench Guidance From Human Demonstration","Dexterous robot manipulation can leverage human demonstrations, yet transferring them into reliable robot policies remains difficult. CHORD introduces Contact Wrench Guidance from Human Demonstration for long-horizon reinforcement learning with rigid and articulated objects. The method uses an object-centric contact wrench space to compare human and robot motions by induced force–torque effects on the object. A large simulation benchmark with 4,739 bimanual tasks enables scalable training and shows 82.12% success on 1,831 tasks, with 90.77% whole-body generalization and real-world transfer.","Learning Dexterous Manipulation Using Contact Wrench Guidance From Human Demonstration  \nXinghao Zhu* ,‡, Zixi Liu* , Shalin Jain* , Chenran Li†, Milad Noori†, Huihua Zhao, John Welsh, Michael Andres Lin, Wei Liu, Tingwu Wang, Xingye Da, Zhengyi Luo, Vishal Kulkarni, Naema Bhatti, Yuke Zhu, Linxi Fan, Bowen Wen, Danfei Xu, Soha Pouya, Yan Chang‡  \nNVIDIA  \n* Equal Contribution †Core Contributor ‡Project Lead and Corresponding Author  \narXiv :2607 .00033v1 [ cs .RO] 22 Jun 2026  \nFigure 1: CHORD learns dexterous, contact-rich policies from human demonstration through contact wrench-space guided reinforcement learning. Foreground: (a) the framework takes reference hand-object trajectories from human demonstration and (b) learns a robot policy in simulation, which (c) transfers to real-robot execution. CHORD enables manipulation of (d) articulated objects, (e) rigid objects, and generalizes to (f, g) whole-body embodiments. Background shows CHORD on our large-scale benchmark ofbimanual manipulation tasks.  \nAbstract: Dexterous robot manipulation can benefit from the abundance of human demonstrations, but transferring such demonstrations to robot policies remains challenging. We present Contact Wrench Guidance from Human Demonstration in Robotic Dexterous Manipulation (CHORD), a framework for long-horizon manipulation of rigid and articulated objects with reinforcement learning. The key idea is object-centric contact wrench space guidance: we represent human and robot motions by the forces and torques they can induce on the object, enabling similarity to be measured by the induced instantaneous motions. This guidance makes reinforcement learning more scalable for contact-rich dexterous manipulation. We further introduce a large-scale simulation benchmark with 4,739 bimanual dexterous manipulation tasks, constructed from motion-capture datasets and reconstructed in-house videos. Evaluated on 1,831 benchmark tasks, CHORD achieves an average success rate of 82 . 12%, demonstrating strong scalability. CHORD also generalizes to whole-body manipulation from hand-only and third-person demonstrations, achieving a 90.77% success rate, and the learned policies transfer to the real world in both open-loop and closed-loop settings. Videos and code are available at [https://nvidia-isaac.github.io/video_to_data/chord/](https://nvidia-isaac.github.io/video_to_data/chord/) .  \n1. Introduction  \nRobotic dexterous hands offer a compelling platform for general-purpose manipulation because of their morphological similarity to human hands, enabling robots to exhibit human-level dexterity. Leveraging human demonstrations has shown promise in advancing dexterous manipulation by mitigating the exploration challenges of policy optimization [15, 18 , 19 , 29 , 49 , 50] and supporting the learning of transferable human-robot representations [9, 11 , 31 , 43 , 51] . However, existing methods that use human demonstrations for optimization often rely on brittle assumptions for how the demonstration should be transferred, while representation-learning approaches typically require aligned human-robot data for each task and object, limiting their scalability beyond curated settings. This motivates our work: How can a robot learn transferable and scalable task knowledge from human demonstrations?  \nA dexterous robot cannot simply replay human hand motion to reproduce human manipulation behavior [49– 51] . Differences in morphology, kinematics, and hand geometries require the robot to use different motionsand contacts to achieve the same functional effect on the object. Our key insight is that contact provides a natural bridge between human demonstrations and robot actions. Rather than matching 3D human contact locations [15, 29 , 49], the robot should find contacts that induce object motions aligned with the human demonstration. To this end, we represent contacts in an object-centric wrench space, capturing the force-torque directions and object motions supported ","cbCail25SVxSeksP","https://ap.wps.com/l/cbCail25SVxSeksP","pdf",41769016,4,1,24,"English","en",105,"# Introduction\n# Related Work","[{\"question\":\"What problem does CHORD address in dexterous robotic manipulation?\",\"answer\":\"Transferring human demonstrations into robot policies is challenging for long-horizon, contact-rich manipulation. CHORD aims to learn transferable, scalable task knowledge without brittle transfer assumptions.\"},{\"question\":\"How does CHORD represent guidance from demonstrations?\",\"answer\":\"CHORD uses object-centric contact wrench space guidance, representing contacts by the force-torque directions and object motions they can induce. Similarity is measured by the induced instantaneous motions.\"},{\"question\":\"What results does CHORD achieve on its benchmark and in real-world transfer?\",\"answer\":\"On 1,831 benchmark tasks, CHORD achieves an average success rate of 82.12%. It generalizes to whole-body manipulation with 90.77% success and transfers policies to real robots in open-loop and closed-loop settings.\"}]",1784175737,60,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"learning-dexterous-manipulation-using-contact-wrench-guidance-from-human-demonstration","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":20},"https://docshare.wps.com/document/learning-dexterous-manipulation-using-contact-wrench-guidance-from-human-demonstration/81737/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-25","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does CHORD address in dexterous robotic manipulation?","Question",{"text":75,"@type":76},"Transferring human demonstrations into robot policies is challenging for long-horizon, contact-rich manipulation. CHORD aims to learn transferable, scalable task knowledge without brittle transfer assumptions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does CHORD represent guidance from demonstrations?",{"text":80,"@type":76},"CHORD uses object-centric contact wrench space guidance, representing contacts by the force-torque directions and object motions they can induce. Similarity is measured by the induced instantaneous motions.",{"name":82,"@type":73,"acceptedAnswer":83},"What results does CHORD achieve on its benchmark and in real-world transfer?",{"text":84,"@type":76},"On 1,831 benchmark tasks, CHORD achieves an average success rate of 82.12%. It generalizes to whole-body manipulation with 90.77% success and transfers policies to real robots in open-loop and closed-loop settings.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,109,114,119,122,127,130,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":29,"slug":108},5,"Comic","comic",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]