[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-141362-105":59,"doc-detail-141362-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","redsdf-regularized-deep-signed-distance-fields-for-robotics","ReDSDF - Regularized Deep Signed Distance Fields for Robotics","","Autonomous robots must avoid unintended contact with obstacles and humans, making distance-based constraints essential for safe planning and control. Different tasks require different distance resolution, motivating heuristic distance-field measurement methods. This paper introduces ReDSDF, a single neural implicit function that computes smooth signed distance fields at multiple scales, delivering fine-grained resolution for high-dimensional manifolds and articulated bodies. Results on simulated whole-body control and safe human-robot interaction demonstrate effectiveness.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/redsdf-regularized-deep-signed-distance-fields-for-robotics/141362/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/redsdf-regularized-deep-signed-distance-fields-for-robotics/141362.png","ImageObject",300,407,{"name":92,"@type":93},"Mali","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-09-26","2026-08-25",true,{"@type":102,"interactionType":103,"userInteractionCount":29},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What problem does ReDSDF address in autonomous robotics?","Question",{"text":112,"@type":113},"It addresses the need for accurate, smooth distance fields to prevent unintended contact between robots and dynamic or complex environments, including humans and other articulated objects.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does ReDSDF compute distance fields across different scales?",{"text":117,"@type":113},"It uses a single neural implicit function with an inductive bias that switches between a learned distance approximation and an L2-regularized distance based on a soft threshold.",{"name":119,"@type":110,"acceptedAnswer":120},"What kinds of systems and tasks are evaluated to demonstrate effectiveness?",{"text":121,"@type":113},"The method is evaluated in representative simulated tasks, including whole-body control and safe human-robot interaction in shared workspaces.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},141362,1787654920,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":29,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":81,"language":139,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":67,"update_tm":129,"read_time":39},2336475104362,"https://ap-avatar.wpscdn.com/avatar/22000c4c46a41b752dd?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786595829695023868","ReDSDF: Regularized Deep Signed Distance Fields for Robotics  \nPuze Liu, Kuo Zhang, Davide Tateo, Snehal Jauhri, Jan Peters and Georgia Chalvatzaki  \nAbstract—The safe operation of autonomous robots requires avoiding unintentional contact between the robot itself and the environment, such as obstacles and humans. Distance-based constraints are fundamental for enabling robots to plan and act safely. Moreover, different applications require different distance resolutions, leading to various heuristic approaches for measuring distance fields w.r.t. obstacles. We propose Regularized Deep Signed Distance Fields (ReDSDF), a single neural implicit function that can compute smooth distance fields at any scale, with fine-grained resolution over high-dimensional manifolds and articulated bodies like humans, thanks to our effective data generation and a simple inductive bias during training. We demonstrate the effectiveness of our approach in representative simulated tasks for whole-body control (WBC) and safe Human-Robot Interaction (HRI) in shared workspaces.  \nI. INTRODUCTION  \nDistance-based constraints restrict robot operation space to avoid unintentional contacts with the environment. Defining distances between query points and objects with complex shapes or dynamic articulations, such as shelves or humans, is a challenging problem. A Signed Distance Function (SDF) constructs an implicit function that inputs the query point and the surface encoding and outputs the signed distance between the query point and the surface. The constraint surface is defined as the 0-level curve. Furthermore, SDFscan be differentiable, enabling the exploitation of gradientbased methods to deal with the constraints.  \nSDFs generally can be categorized into sensor-centric SDFand object-centric SDF. Sensor-centric SDF tries to reconstruct the partially observed surface of the environment based on the distance information obtained from the sensor. Peer research focuses on the improvement of surface reconstruction, such as robustness [1], refinement [2] . Sensor-centric SDF are widely used for mobile robot mapping, planning, and navigation [3]–[9] . However, sensor-centric SDF are not applicable when the query point differs from the sensor. Object-centric SDF approximates the object surface as alevel-0 curve. We can query the distances of any points in the space w.r.t the object reference frame. Object-centric SDF can be approximated by simple geometry primitives, such as spheres and capsules [10] or deep neural networks [11],[12] . Object-centric approaches majorly focus on object surface reconstruction and are beneficial in robotic tasks, such as  \nTechnische Universitt Darmstadt, Karolinenpl. 5, 64289 Darmstadt, Germany [puze@robot-learning.de](puze@robot-learning.de) ,  \n[kuo.zhang@stud.tu-darmstadt.de](kuo.zhang@stud.tu-darmstadt.de) , {davide.tateo, snehal.jauhri, jan.peters, georgia.chalvatzaki}@[tu-darmstadt.de](tu-darmstadt.de)  \nThis research received funding by the DFG Emmy Noether Programme (\\#448644653), the RoboTrust project of the Centre Responsible Digitality Hessen, Germany and the China Scholarship Council (No. 201908080039) .  \nmanipulation [13]–[16] . However, object-centric approaches majorly focus on object surface reconstruction but are lacking accuracy outside of the object boundary. The incorrect distance information may provide a misleading sense of safety and cause hazardous behavior. In addition, existing approaches do not consider the articulated objects, while our proposed approach tries to fill this gap.  \nIn this paper, we present ReDSDF, an approach that extends the concept of SDFs for arbitrary articulated objects such as robotic manipulators and humans. Our method provides a single deep model that learns distance fields which can be used at any scale. Notably, we define a simple yet effective inductive bias that regularizes implicit function to approximate the distance by the L2 norm w.r.t. the object’s center. This regular","cbCaiipvKH0bu6t3","https://ap.wps.com/l/cbCaiipvKH0bu6t3","pdf",563530,"English","# Introduction\n## Signed Distance Functions and Their Limitations\n# Learning ReDSDF for Robot Control\n## Model Definition\n## Data Generation and Training","[{\"question\":\"What problem does ReDSDF address in autonomous robotics?\",\"answer\":\"It addresses the need for accurate, smooth distance fields to prevent unintended contact between robots and dynamic or complex environments, including humans and other articulated objects.\"},{\"question\":\"How does ReDSDF compute distance fields across different scales?\",\"answer\":\"It uses a single neural implicit function with an inductive bias that switches between a learned distance approximation and an L2-regularized distance based on a soft threshold.\"},{\"question\":\"What kinds of systems and tasks are evaluated to demonstrate effectiveness?\",\"answer\":\"The method is evaluated in representative simulated tasks, including whole-body control and safe human-robot interaction in shared workspaces.\"}]","ReDSDF - Regularized Deep Signed Distance Fields for Robotics | PDF"]