[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-system-learns-3d-object-relationships-without-labels":10,"sections":34},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},9691,"ai-system-learns-3d-object-relationships-without-labels","AI System Learns 3D Object Relationships Without Labels","A new feed-forward model finds how objects relate in 3D scenes across multiple camera views, without needing object labels or known camera positions.","Researchers have built an AI system that figures out how objects relate to each other in 3D space, not just what the objects are.\n\nThe model, called RelationVGGT, takes multiple camera views of a scene plus a text query like \"the cup next to the lamp\" and highlights the matching object across every view. It does this without ever being told the object's category name, and without needing camera positions calculated in advance - a step that normally requires per-scene optimization. The system combines semantic features from an existing visual foundation model with geometry-aware representations from a 3D geometry foundation model, then feeds both into a relation transformer that links the visual subject to the text query. The team also built an automated pipeline on the ScanNet++ dataset, using vision-language models and large language models to generate training annotations at scale, since no existing dataset labels spatial relations this way.\n\nMost 3D scene-understanding systems can tell you an object is a \"chair\" or a \"table,\" but not that the chair is behind the table. That gap matters for anything that has to navigate or manipulate real spaces - robots, AR headsets, inspection tools - where \"behind,\" \"next to,\" and \"on top of\" are the actual instructions, not labels.\n\nIt's a research paper, not a product, and its training pipeline leans on other AI models to grade its own homework - worth remembering before anyone calls this solved.","[\"ai\",\"computer-vision\",\"3d-vision\",\"research\"]","2026-10-02T04:00:00.000Z","2026-10-03T06:33:46.971Z","2026-10-03T06:33:52.356Z","published",null,[],"ai",[24,26,27,28],"computer-vision","3d-vision","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.00970",0,{"sections":35},[36,39,43,47,52,56,60,65,70,75,80,85,90,95],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",5976,{"name":40,"slug":41,"count":42,"latest_published_at":18},"Security","security",842,{"name":44,"slug":45,"count":46,"latest_published_at":18},"Policy","policy",438,{"name":48,"slug":49,"count":50,"latest_published_at":51},"Deals","deals",317,"2026-10-01T22:00:00.000Z",{"name":53,"slug":54,"count":55,"latest_published_at":18},"Hardware","hardware",199,{"name":57,"slug":58,"count":59,"latest_published_at":18},"Science","science",173,{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",155,"2026-10-01T19:54:10.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Dev Tools","dev-tools",96,"2026-10-01T16:57:03.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Software","software",93,"2026-09-30T21:41:11.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Startups","startups",90,"2026-10-01T21:55:22.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",53,"2026-10-02T02:50:39.000Z",{"name":86,"slug":87,"count":88,"latest_published_at":89},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":91,"slug":92,"count":93,"latest_published_at":94},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":96,"slug":97,"count":98,"latest_published_at":99},"How-To","how-to",7,"2026-10-01T09:00:00.000Z"]