[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85532-en":3,"doc-seo-85532-105":28,"detail-sidebar-cat-0-en-105":90},{"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":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":13,"seo_description":14,"update_tm":26,"read_time":27},85532,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","RVN-Bench: A Benchmark for Reactive Visual Navigation","Safe visual navigation is critical for indoor mobile robots in cluttered spaces, yet many existing benchmarks omit collision handling or target outdoor driving, limiting their suitability for indoor reactive navigation. RVN-Bench introduces a collision-aware benchmark where an agent reaches sequential goal positions in previously unseen environments using only visual observations and no prior map. Built on Habitat 2.0 with high-fidelity HM3D scenes, it provides task definitions, evaluation metrics, standardized training tools, and online/offline learning support. Results show effective generalization across unseen simulated environments and promising sim-to-real transfer from initial Jackal UGV experiments.","RVN-Bench: A Benchmark for Reactive Visual Navigation  \nJaewon Lee 1 , Jaeseok Heo2 , Gunmin Lee2 , Howoong Jun3 , Jeongwoo Oh4 and Songhwai Oh 1  \narXiv :2603 .03953v2 [ cs .RO] 11 Jul 2026  \nAbstract—Safe visual navigation is critical for indoor mobile robots operating in cluttered environments. Existing benchmarks, however, often neglect collisions or are designed for outdoor scenarios, making them unsuitable for indoor visual navigation. To address this limitation, we introduce the reactive visual navigation benchmark (RVN-Bench), a collisionaware benchmark for indoor mobile robots. In RVN-Bench, an agent must reach sequential goal positions in previously unseen environments using only visual observations and no prior map, while avoiding collisions. Built on the Habitat 2.0 simulator and leveraging high-fidelity HM3D scenes, RVN-Bench provides large-scale, diverse indoor environments, defines a collision-aware navigation task and evaluation metrics, and offers tools for standardized training and benchmarking. RVN-Bench supports both online and offline learning by offering an environment for online reinforcement learning, a trajectory image dataset generator, and tools for producing negative trajectory image datasets that capture collision events. Evaluations demonstrate that policies trained on RVN-Bench generalize effectively across unseen simulated environments. Furthermore, initial physical experiments using a Jackal UGV indicate promising sim-to-real transfer. Code and additional materials are available at: [https://sequor-robotics-research](https://sequor-robotics-research) . [github.io/projects/RVN-Bench/](github.io/projects/RVN-Bench/).  \nI. INTRODUCTION  \nReactive visual navigation (RVN) is the problem of reaching specified goals while avoiding collisions with obstacles in previously unseen environments using only visual observations, without relying on a prior map or task-specific knowledge. RVN is critical for autonomous mobile robots, which must maintain safety under unexpected environmental changes. Despite recent advances in visual navigation foundation models [1]–[4] that have shown promising performance on reactive visual navigation tasks, the problem of ensuring safety, particularly in obstacle-rich indoor settings where collisions are likely, remains unsolved. Moreover, these approaches typically require both a massive amount of training data and reliable mechanisms for safe evaluation.  \n1J. Lee, J. Heo, G. Lee and S. Oh are with the Department of Electrical and Computer Engineering, Seoul National University (SNU) and Automation and Systems Research Institute (ASRI) and Sequor Robotics Inc., Seoul, Korea (Republic of). [jaewon.lee@rllab.snu.ac.kr](jaewon.lee@rllab.snu.ac.kr) , [jaeseok.heo@rllab.snu.ac.kr](jaeseok.heo@rllab.snu.ac.kr) , [songhwai@snu.ac.kr](songhwai@snu.ac.kr)  \n2 Gunmin Lee is with the Department of Electrical and Computer Engineering, Seoul National University (SNU) and Automation and Systems Research Institute (ASRI), Seoul, Korea (Republic of) . [gunmin.lee@rllab.snu.ac.kr](gunmin.lee@rllab.snu.ac.kr)  \n3Howoong Jun is with Interdisciplinary Program in Artificial Intelligence, Seoul National University (SNU) and Automation and Systems Research Institute (ASRI) and Sequor Robotics Inc., Seoul, Korea (Republic of) . [howoong.jun@rllab.snu.ac.kr](howoong.jun@rllab.snu.ac.kr)  \n3Jeongwoo Oh is with Sequor Robotics Inc., Seoul, Korea (Republic of) . [jeongwoo.oh@sequorrobotics.com](jeongwoo.oh@sequorrobotics.com)  \nCorresponding author: Songhwai Oh  \nCollecting such data directly in the real world is costly and time-consuming, while real-world evaluation is often unsafe, as unverified algorithms risk damaging property or degrading hardware. These challenges motivate the development of simulation frameworks that support scalable training and collision-aware evaluation of RVN.  \nTo address these challenges, a wide range of simulation environments have been utilized to support the design and","cbCaid5azLWFzxdL","https://ap.wps.com/l/cbCaid5azLWFzxdL","pdf",9543836,1,"English","en",105,"# Introduction\n## Reactive visual navigation problem\n## Limitations of existing benchmarks\n## Motivation for a collision-aware indoor benchmark\n## Overview of RVN-Bench","[{\"question\":\"What problem does RVN-Bench target?\",\"answer\":\"RVN-Bench targets reactive visual navigation for indoor mobile robots that must reach sequential goals while avoiding collisions in previously unseen environments using only visual observations and no prior map.\"},{\"question\":\"How does RVN-Bench differ from existing navigation benchmarks?\",\"answer\":\"Many benchmarks either focus on outdoor/autonomous driving scenarios or evaluate only goal reaching while ignoring collisions. RVN-Bench is designed specifically for indoor navigation and treats collision avoidance as a primary evaluation criterion.\"},{\"question\":\"What tools and learning modes does RVN-Bench provide?\",\"answer\":\"RVN-Bench is built on Habitat 2.0 and uses HM3D scenes, offering an environment for online reinforcement learning plus dataset tools, including trajectory image generation and negative trajectory datasets that capture collision events.\"}]",1784204232,20,{"code":4,"msg":29,"data":30},"ok",{"site_id":23,"language":22,"slug":31,"title":13,"keywords":32,"description":14,"schema_data":33,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":26},"rvn-bench-a-benchmark-for-reactive-visual-navigation","",{"@graph":34,"@context":84},[35,52,67],{"@type":36,"itemListElement":37},"BreadcrumbList",[38,42,46,49],{"item":39,"name":40,"@type":41,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":43,"name":44,"@type":41,"position":45},"https://docshare.wps.com/document/","Document",2,{"item":47,"name":12,"@type":41,"position":48},"https://docshare.wps.com/document/research-report/",3,{"item":50,"name":13,"@type":41,"position":51},"https://docshare.wps.com/document/rvn-bench-a-benchmark-for-reactive-visual-navigation/85532/",4,{"url":50,"name":13,"@type":53,"author":54,"headline":13,"publisher":56,"fileFormat":59,"inLanguage":22,"description":14,"dateModified":60,"datePublished":61,"encodingFormat":59,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":55},"Person",{"url":39,"name":57,"@type":58},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":20},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does RVN-Bench target?","Question",{"text":74,"@type":75},"RVN-Bench targets reactive visual navigation for indoor mobile robots that must reach sequential goals while avoiding collisions in previously unseen environments using only visual observations and no prior map.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does RVN-Bench differ from existing navigation benchmarks?",{"text":79,"@type":75},"Many benchmarks either focus on outdoor/autonomous driving scenarios or evaluate only goal reaching while ignoring collisions. RVN-Bench is designed specifically for indoor navigation and treats collision avoidance as a primary evaluation criterion.",{"name":81,"@type":72,"acceptedAnswer":82},"What tools and learning modes does RVN-Bench provide?",{"text":83,"@type":75},"RVN-Bench is built on Habitat 2.0 and uses HM3D scenes, offering an environment for online reinforcement learning plus dataset tools, including trajectory image generation and negative trajectory datasets that capture collision events.","https://schema.org",{"og:url":50,"og:type":86,"og:title":13,"og:site_name":57,"og:description":14},"article",{"robots":88,"canonical":50},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":44,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":45,"doc_module":4,"doc_module_name":44,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":51,"doc_module":4,"doc_module_name":44,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":44,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":44,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":44,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":44,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":44,"category_name":124,"show_sort_weight":27,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":27,"doc_module":4,"doc_module_name":44,"category_name":127,"show_sort_weight":27,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":44,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":44,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]