[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84857-en":3,"doc-seo-84857-105":29,"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":4,"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},84857,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","TRIG: Trajectory-Rig Decoupled Metric Geometry Learning","TRIG (Trajectory-Rig Decoupled Metric Geometry Learning) targets vision-centric autonomous driving by improving metric scene geometry and ego-motion estimation from synchronized multi-camera observations under rigid camera rigs. The method factorizes camera poses into a time-varying ego trajectory and a static camera-rig component, enabling separate modeling and supervision of motion dynamics versus rig topology. Sparse Temporal–Spatial Attention separates cross-camera interaction from temporal aggregation, reducing global attention cost while preserving geometric reasoning. Experiments on five benchmarks achieve state-of-the-art results in pose estimation, metric depth prediction, and 3D reconstruction.","TRIG: Trajectory-Rig Decoupled Metric Geometry Learning  \nLizhou Liao 1 ,†,∗ , Wentao Xu 1 ,2 ,†, Handong Wang 1 , Lirong Yang 1 ,‡, Shuai Yang 1 , Weiwei Liu 1 , Chang Huang 1  \narXiv :2607 .0580 1v 1 [ cs .CV] 7 Jul 2026  \nAbstract—Vision-centric autonomous driving requires accurate metric geometry and ego-motion estimation from synchronized multi-camera observations. Recent visual geometry models show strong performance in pose estimation, depth prediction, and 3D reconstruction, but are not tailored to rigid multi-camera driving systems. They often encode camera poses as entangled representations, in which time-varying ego-motion and static camera-rig geometry are jointly modeled, limiting the utilization of vehicle-side geometric priors. We propose Trajectory-Rig Decoupled Metric Geometry Learning (TRIG), a geometry perception framework for autonomous driving. TRIG factorizes camera poses into ego-trajectory and camerarig components, enabling separate modeling of ego-motion and static multi-camera topology. We introduce decoupled pose encoding and supervision, which separately constrain trajectory evolution and rig geometry for metric-consistent learning. Moreover, sparse Temporal–Spatial attention separates crosscamera interaction from temporal aggregation, reducing global attention cost while preserving geometric reasoning. Experiments on five autonomous driving benchmarks show that TRIG achieves state-of-the-art performance in pose estimation, metric depth prediction, and 3D reconstruction.  \nI. INTRODUCTION  \nVision-centric autonomous driving has emerged as a scalable and cost-effective paradigm for perception, thereby garnering escalating scholarly and industrial interest [1]–[5] . A fundamental challenge in this field is recovering accurate metric scene geometry and ego-motion from synchronized multi-camera observations under rigid camera configurations. Accurate geometric understanding is essential for downstream applications such as planning [2], simulation [6], and online HD map construction [7] . Recent visual geometry models have demonstrated impressive capabilities in camera pose estimation, depth prediction, and 3D reconstruction from image sequences [8]–[10] . However, directly applying these models to autonomous driving remains challenging due to rigid multi-camera configurations and the requirement for accurate metric geometry.  \nDespite substantial progress, reliable metric geometry estimation remains challenging for multi-camera autonomous driving. Existing methods are largely image-driven, using vehicle-side geometry, such as odometry and camera-rig calibration, only as auxiliary inputs or supervision rather than explicit representations for reasoning [8], [10]–[12] .  \n†These authors contributed equally to this work.  \n1Carizon. E-mail: {lizhou .liao, intern .wentao .xu, [handong.wang](handong.wang) , lirong.yang, shuai.yang, weiwei.liu, [chang.huang](chang.huang}@carizon.com)[}](chang.huang}@carizon.com)[@carizon.com](chang.huang}@carizon.com).  \n2 ShanghaiTech University. All work was done during the internship.  \n∗Corresponding author.  \n‡Project leader.  \nFig. 1: Radar chart of method rankings across five benchmarks. Radial axes represent datasets; rank scores (0– 1, higher is better) are averaged over five sub-metrics (Acc, Comp, AbsRel, δ1.25 , AUC@30◦ ) .  \nAs a result, metric scale is implicitly inferred from visual appearance, causing scale ambiguity and limited geometric accuracy. Recent geometry-prior methods [13]–[15] are mainly designed for generic visual scenes and treat images as independent observations, thus under-exploiting the structural constraints of synchronized rigid camera rigs. A camera pose in autonomous driving naturally consists of a time-varying ego trajectory and a static camera-rig configuration. However, existing methods typically entangle them into a unified pose representation, making it difficult to separately exploit motion dynamics and camera topology. Such entangle","cbCaiaLxpoAsNS6u","https://ap.wps.com/l/cbCaiaLxpoAsNS6u","pdf",4856517,1,9,"English","en",105,"# Introduction\n## Problem: metric geometry and ego-motion under rigid multi-camera rigs\n## Limitations of entangled pose representations\n# Proposed Method\n## Trajectory-Rig decoupled framework (TRIG)\n## Decoupled pose encoding and supervision\n## Sparse Temporal–Spatial Attention (STSA)","[{\"question\":\"What core idea does TRIG introduce for multi-camera autonomous driving?\",\"answer\":\"TRIG decouples camera pose into a time-varying ego trajectory and a static camera-rig configuration. This separates motion dynamics from rig topology for more accurate metric learning.\"},{\"question\":\"Why are entangled pose representations considered problematic in this setting?\",\"answer\":\"Entanglement jointly models ego-motion and static rig geometry, limiting the use of vehicle-side geometric priors. It also forces expensive global reasoning such as Sim(3) alignment to achieve consistent metric-scale reconstruction.\"},{\"question\":\"How does Sparse Temporal–Spatial Attention (STSA) help TRIG?\",\"answer\":\"STSA performs continuous temporal trajectory aggregation while injecting rig constraints sparsely across cameras. This separates cross-camera interaction from temporal aggregation, lowering global attention cost while maintaining geometric reasoning.\"}]",1784198852,23,{"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":85,"head_meta":87,"extra_data":89,"updated_unix":27},"trig-trajectory-rig-decoupled-metric-geometry-learning","",{"@graph":35,"@context":84},[36,53,67],{"@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/trig-trajectory-rig-decoupled-metric-geometry-learning/84857/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What core idea does TRIG introduce for multi-camera autonomous driving?","Question",{"text":74,"@type":75},"TRIG decouples camera pose into a time-varying ego trajectory and a static camera-rig configuration. This separates motion dynamics from rig topology for more accurate metric learning.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"Why are entangled pose representations considered problematic in this setting?",{"text":79,"@type":75},"Entanglement jointly models ego-motion and static rig geometry, limiting the use of vehicle-side geometric priors. It also forces expensive global reasoning such as Sim(3) alignment to achieve consistent metric-scale reconstruction.",{"name":81,"@type":72,"acceptedAnswer":82},"How does Sparse Temporal–Spatial Attention (STSA) help TRIG?",{"text":83,"@type":75},"STSA performs continuous temporal trajectory aggregation while injecting rig constraints sparsely across cameras. This separates cross-camera interaction from temporal aggregation, lowering global attention cost while maintaining geometric reasoning.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"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":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":123,"show_sort_weight":124,"slug":125},"Religion & Spirituality",20,"religion-spirituality",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":124,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]