[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83974-en":3,"doc-seo-83974-105":29,"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":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},83974,7971461740886,"Theodore","https://ap-avatar.wpscdn.com/davatar_3d24733baf745e90a7e4bdd5f77d97b2",8,"Research & Report","From Visual Geometry Evidence to Embodied Semantic Occupancy Memory","Semantic occupancy provides a structured spatial memory for embodied indoor agents by jointly representing occupied regions, observed free space, unknown areas, and object semantics. Existing indoor benchmarks mainly address single-view prediction or room-level online perception, leaving long-horizon semantic mapping across connected indoor spaces less studied. This work introduces HIOcc, a hierarchical benchmark unifying ScanNet, ScanNet++, and Matterport3D. It preserves native observation geometries and supports local prediction, room-level mapping, and building-level panoramic mapping.","arXiv :2607 .05543v 1 [ cs .RO] 6 Jul 2026  \nFrom Visual Geometry Evidence to Embodied Semantic Occupancy Memory  \nHu Zhu 12 Bohan Li23 Xianda Guo4 Hongsi Liu2 Baorui Peng6 Mingqi Yuan5 Xin Jin2 Wenjun Zeng2∗ Chang Wen Chen 1  \n1The Hong Kong Polytechnic University 2Eastern Institute of Technology 3 Shanghai Jiao Tong University  \n4Wuhan University 5The University of Hong Kong 6 Georgia Institute of Technology  \nAbstract: Semantic occupancy provides a structured spatial memory for embodied indoor agents by jointly representing occupied regions, observed free space, unknown areas, and object semantics. However, existing indoor occupancy benchmarks and methods mainly focus on single-view prediction or room-level online perception, leaving long-horizon semantic mapping across connected indoor spaces underexplored. We introduce HIOcc, a hierarchical indoor occupancy benchmark that unifies ScanNet, ScanNet++, and Matterport3D under a common sparse semantic occupancy format while preserving their native observation geometries, including perspective RGB-D frames and pano-centric observation groups. HIOcc supports three complementary evaluation regimes: local semantic occupancy prediction, room-level online occupancy mapping, and building-level mapping across connected panoramic environments.  \nWe further propose GEM-Occ, a Gaussian Evidence Memory framework for semantic occupancy mapping. Rather than using pointmaps as persistent map states, GEM-Occ treats local visual geometry predictions as transient evidence, converts them into semantic Gaussian occupancy evidence and free-space ray evidence, and fuses them into a persistent hierarchical memory through visibilityand uncertainty-aware causal updates. The memory is organized into local caches, room-level submaps, and a building-level graph, and can be queried at any time through Gaussian-to-occupancy splatting. Experiments on HIOcc show that GEM-Occ improves local occupancy prediction, online map stability, free-space reasoning, revisit consistency, and building-level scalability over prior indoor occupancy and Gaussian-based mapping baselines.  \nKeywords: Semantic Occupancy, Streaming Map, Robotics Perception  \n1 Introduction  \nAs embodied agents move through indoor environments, 3D perception becomes a memory problem: the agent must accumulate egocentric observations into a persistent, queryable representation of occupied structures, observed free space, unknown regions, and semantic categories. Semantic occupancy naturally supports this setting by jointly encoding geometry and semantics for navigation, exploration, interaction, and downstream scene understanding. Early indoor semantic scene completion and monocular occupancy methods mainly predict local volumes from partial observations [1–7], while outdoor autonomous-driving works have established occupancy as a scalable representation for surround-view perception, forecasting, and scene generation [8–23] . Recent 3D Gaussian Splatting [24] based methods further replace dense voxel queries with compact continuous primitives and Gaussian-to-voxel splatting [25–30] . Despite this progress, most formulations remain single-frame, local, offline, or bounded by a predefined room-scale volume, leaving long-horizon embodied mapping underexplored, where the map must be updated causally, distinguish free space from unknown space, remain stable under revisits, and scale from local views to rooms and connected buildings.  \n*  \nCorresponding author  \nFigure 1: Overview of HIOcc and GEM-Occ. HIOcc provides hierarchical indoor semantic occupancy annotations across local views, rooms, and connected buildings. GEM-Occ builds persistent semantic Gaussian memory for local prediction, room-level online mapping, and buildinglevel mapping.  \nExisting benchmarks reflect the same limitation. Foundational indoor datasets such as NYUv2, ScanNet, ScanNet++, and Matterport3D have enabled substantial progress in RGB-D perception and 3D scene understanding [","cbCainradrgzPfwW","https://ap.wps.com/l/cbCainradrgzPfwW","pdf",15664839,1,19,"English","en",105,"# Introduction\n## Semantic occupancy as a memory problem\n## Limitations of existing indoor benchmarks\n## HIOcc: hierarchical occupancy benchmark\n## GEM-Occ: Gaussian evidence memory","[{\"question\":\"What does semantic occupancy mean for embodied indoor agents?\",\"answer\":\"Semantic occupancy is a structured spatial memory that jointly encodes occupied regions, observed free space, unknown areas, and semantic categories for navigation and scene understanding.\"},{\"question\":\"Why are existing indoor occupancy benchmarks insufficient for long-horizon mapping?\",\"answer\":\"They often focus on single-view prediction or room-level online perception, without fully supporting causal long-horizon semantic mapping across connected spaces.\"},{\"question\":\"How does GEM-Occ build persistent memory from visual predictions?\",\"answer\":\"GEM-Occ treats local visual geometry predictions as transient evidence, converts them into semantic Gaussian occupancy evidence and free-space ray evidence, and fuses them into a hierarchical memory using visibility and uncertainty-aware causal updates.\"}]",1784191790,48,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"from-visual-geometry-evidence-to-embodied-semantic-occupancy-memory","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/from-visual-geometry-evidence-to-embodied-semantic-occupancy-memory/83974/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","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 does semantic occupancy mean for embodied indoor agents?","Question",{"text":75,"@type":76},"Semantic occupancy is a structured spatial memory that jointly encodes occupied regions, observed free space, unknown areas, and semantic categories for navigation and scene understanding.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why are existing indoor occupancy benchmarks insufficient for long-horizon mapping?",{"text":80,"@type":76},"They often focus on single-view prediction or room-level online perception, without fully supporting causal long-horizon semantic mapping across connected spaces.",{"name":82,"@type":73,"acceptedAnswer":83},"How does GEM-Occ build persistent memory from visual predictions?",{"text":84,"@type":76},"GEM-Occ treats local visual geometry predictions as transient evidence, converts them into semantic Gaussian occupancy evidence and free-space ray evidence, and fuses them into a hierarchical memory using visibility and uncertainty-aware causal updates.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},"General","general"]