[{"data":1,"prerenderedAt":201},["ShallowReactive",2],{"method-affan2026semanticmeshing":3},{"method":4,"reference":53,"equipment":73,"figures":98,"results":139},{"id":5,"label":6,"shortName":7,"title":8,"year":9,"era":10,"cluster":11,"scope":12,"keyIdeaZh":13,"keyIdeaEn":14,"fulltextStatus":15,"publicationStatus":16,"recommendation":17,"constructionRelevance":18,"validationEnvironment":19,"strengths":23,"limitations":25,"sensors":28,"platform":32,"estimator":36,"association":37,"timeModel":38,"deskew":39,"loopClosure":40,"globalOptimization":41,"mapRepresentation":42,"prior":43,"outputGeometry":44,"compute":45,"codeUrl":46,"codeLicense":47,"relatedVersions":48},"affan2026semanticmeshing","Affan et al., 2026","Semantics-aided incremental meshing (LIO + RGB)","Incremental Semantics-Aided Meshing from LiDAR-Inertial Odometry and RGB Direct Label Transfer",2026,"recent","C12","map_representation_or_reconstruction","此方法以 OneFormer 視覺基礎模型對每張 RGB 影像做全景分割，再利用 FAST-LIO2 的 IMU 狀態把標籤投影到已完成運動畸變校正（deskew）的 LiDAR 掃描點，並以遮罩侵蝕、邊界距離與深度不連續檢查剔除不可靠的投影。帶標籤的點進入改良的 TSDF：每個體素保存標籤直方圖，截斷距離依類別與量測距離調整（例如欄杆較窄、牆面較寬），最後以 Marching Cubes 產生帶語意的網格，可匯出為 USD 資產。作者刻意不用迴圈閉合，以免位姿跳動使增量式 TSDF 失效。","Projects OneFormer panoptic labels from RGB frames onto FAST-LIO2-deskewed LiDAR scans and fuses them into a Voxblox-style TSDF with per-voxel label histograms and class- and range-dependent truncation, then extracts a labelled marching-cubes mesh exportable to USD; loop closure is deliberately omitted.","full_text_reviewed","status_uncertain","supplementary","定量評估僅用 Oxford Spires 的 Christ Church College（sequence 2，TLS 真值）；作者在 Sec. 5.2 稱其為室內文化資產場景，但 Sec. 5.1.1 的定性描述同時涵蓋室外與室內空間（尖塔與拱形結構）。場景屬完工建築而非施工中工地；NTU VIRAL（NYA01）只作定性分析。引言提及 AEC 與數位分身用途，但論文未做營建場域驗證。",[20,21,22],"public_benchmark","completed_building","independent_reference",[24],"On Oxford Spires Christ Church College (sequence 2) with TLS ground truth: accuracy RMSE 14.54 cm, completeness 98.55% and F1 98.58%, versus Voxblox 17.24 cm\u002F97.17%\u002F96.98% and ImMesh 17.40 cm\u002F95.99%\u002F96.31% (Sec. 5.2, Table 1).",[26,27],"[\"Authors list odometry drift on longer trajectories, dependence on camera field-of-view coverage and calibration, and thin structures, specular materials and heavy clutter (Sec. 6).\", \"Meshes were aligned to the TLS ground truth by manual coarse registration plus ICP before scoring, so the metrics exclude global pose error (Sec. 5.2).\", \"Geometric and semantic uncertainty overlap only slightly, and the authors note that mismatched assumptions between LIO back-end, deskewing and meshing libraries can propagate into the mesh (Sec. 5.3, Table 2).\", \"Class multipliers, the sparsity factor and the weight cap were set empirically (Sec. 3.4).\", \"(observation) The statement that LIO drift stays within voxel-resolution tolerance on the evaluated trajectories (Sec. 3.1) is not backed by a reported trajectory error metric.\", \"(inference) Learned semantic priors can alter reconstructed geometry at boundaries","independence of geometric evidence must be checked for engineering use.\"]",[29,30,31],"3D LiDAR","IMU","monocular camera",[33,34,35],"[\"public multi-sensor datasets only (Oxford Spires Christ Church College sequence 2","NTU VIRAL NYA01)","the carrying platforms are not described in the paper\"]","FAST-LIO2 tightly coupled LiDAR-inertial odometry (Sec. 3.1)","LiDAR registration inherited from FAST-LIO2 (nearest-neighbour search on ikd-Tree, Sec. 3.1); image-to-LiDAR label association by ego-motion-compensated pinhole-plus-distortion projection filtered by mask erosion, boundary-distance rejection and depth-discontinuity checks (Sec. 3.2); voxel labels by majority vote plus greedy segment-to-map-label overlap assignment (Sec. 3.3)","IMU state forward-propagated from the LiDAR timestamp to the camera timestamp to form a dynamic LiDAR-to-camera transform, with per-point motion correction from the instantaneous camera velocity (Sec. 3.2, Eq. 1)","labels are transferred onto point clouds deskewed by FAST-LIO2 (Sec. 3)","none (authors deliberately use LIO without loop closure to avoid discontinuous pose corrections that would require TSDF re-integration, Sec. 3.1)","none","Voxblox-based TSDF with 0.1 m voxels, per-voxel label histogram with range-dependent log-normal confidence, class- and range-dependent truncation, base weighting either constant or inverse-square (1\u002Fz^2 of the point coordinate in the sensor frame; the mode used in the experiments is not stated), behind-surface weight dropoff, sparsity compensation factor 1.5 and weight cap 255, plus per-voxel semantic and geometric uncertainty scores; mesh via marching cubes","vision foundation model labels (OneFormer) on RGB frames (learned prior)","semantically labelled triangle mesh (USD assets mentioned)","not_reported (no processor, GPU or timing figures are given; the pipeline is only stated to operate at LiDAR rate, Sec. 3)",null,"not_verified",[49],{"relation":50,"title":51,"doi_or_url":52},"preprint","arXiv:2604.09478","https:\u002F\u002Farxiv.org\u002Fabs\u002F2604.09478",{"id":5,"kind":54,"shortName":7,"title":8,"authors":55,"year":9,"venue":59,"venueType":60,"publisher":61,"volumeIssuePages":62,"doi":63,"arxivId":64,"url":52,"firstPublicDate":65,"publicationStatus":16,"metadataStatus":66,"fulltextStatus":15,"era":10,"classicReason":67,"codeUrl":46,"cluster":11,"topics":68,"mdpi":69,"verification":70,"label":6,"fulltextRoute":71,"versionRead":72,"addedByCensus":69},"method",[56,57,58],"Muhammad Affan","Ville Lehtola","George Vosselman","The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences","conference","Copernicus","XLIX-B2-2026:335-342","10.5194\u002Fisprs-archives-xlix-b2-2026-335-2026","2604.09478","2026-04-10","metadata_verified","not_applicable",[11],false,"corrected","publisher OA","Version of record, ISPRS Archives XLIX-B2-2026, pp. 335-342 (Copernicus PDF, CC BY 4.0), compared word by word with arXiv 2604.09478v1 (10 Apr 2026, 8 pp.)",[74,81,86,91],{"category":75,"model":76,"canonical":76,"role":77,"dataset":78,"specs":79,"locator":80},"lidar","3D LiDAR (model not reported)","method input","Oxford Spires; NTU VIRAL","maximum range about 150 m stated for the indoor LiDAR setup","Sec. 3; Sec. 5.2.2",{"category":82,"model":83,"canonical":83,"role":77,"dataset":78,"specs":84,"locator":85},"imu","IMU (model not reported)","not_reported","Sec. 3, Sec. 3.2",{"category":87,"model":88,"canonical":88,"role":77,"dataset":78,"specs":89,"locator":90},"camera","RGB camera (model not reported)","pinhole-plus-distortion projection model; calibrated LiDAR-to-camera extrinsics","Sec. 3.2",{"category":92,"model":93,"canonical":93,"role":94,"dataset":95,"specs":96,"locator":97},"tls_scanner","terrestrial laser scanner (model not reported)","reference or ground truth","Oxford Spires","TLS ground truth of the Christ Church College scene","Sec. 5.2",[99,112,122,131],{"refId":5,"refLabel":6,"fig":100,"whatZh":101,"license":102,"licenseUrl":103,"sourceUrl":104,"src":105,"width":106,"height":107,"thumb":108,"thumbWidth":109,"thumbHeight":110,"modified":111},"Fig. 2","系統流程圖：RGB 全景分割、FAST-LIO2 里程計與標籤轉移、語意感知 TSDF 融合與網格化","CC BY 4.0","https:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby\u002F4.0\u002F","https:\u002F\u002Farxiv.org\u002Fhtml\u002F2604.09478v1\u002Ffigures\u002Fvoxblox_augmented_pipeline.png","\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-2.webp",1314,550,"\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-2.thumb.webp",480,201,"converted to WebP",{"refId":5,"refLabel":6,"fig":113,"whatZh":114,"license":102,"licenseUrl":103,"sourceUrl":115,"src":116,"width":117,"height":118,"thumb":119,"thumbWidth":109,"thumbHeight":120,"modified":121},"Fig. 3(c)","本方法在 Oxford Spires（Christ Church College）重建的網格，可與 ImMesh、Voxblox 對照","https:\u002F\u002Farxiv.org\u002Fhtml\u002F2604.09478v1\u002Ffigures\u002Fours_crop_spires_annotated.png","\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-3-c.webp",1400,1065,"\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-3-c.thumb.webp",365,"resized to at most 1400 px wide and converted to WebP",{"refId":5,"refLabel":6,"fig":123,"whatZh":124,"license":102,"licenseUrl":103,"sourceUrl":125,"src":126,"width":127,"height":128,"thumb":129,"thumbWidth":109,"thumbHeight":130,"modified":111},"Fig. 4(c)","本方法在 NTU VIRAL（NYA01）重建的網格，作者標註完整度較高的區域","https:\u002F\u002Farxiv.org\u002Fhtml\u002F2604.09478v1\u002Ffigures\u002Fours_ntuviral_crop_annotated.png","\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-4-c.webp",1228,1112,"\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-4-c.thumb.webp",435,{"refId":5,"refLabel":6,"fig":132,"whatZh":133,"license":102,"licenseUrl":103,"sourceUrl":134,"src":135,"width":117,"height":136,"thumb":137,"thumbWidth":109,"thumbHeight":138,"modified":121},"Fig. 5(a)","NTU VIRAL 重建的幾何不確定度分布，集中於深度不連續與高曲率處","https:\u002F\u002Farxiv.org\u002Fhtml\u002F2604.09478v1\u002Ffigures\u002Fvoxblox_augment_zoompp_geom200.png","\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-5-a.webp",977,"\u002Ffigure-files\u002Faffan2026semanticmeshing\u002Ffig-5-a.thumb.webp",335,{"totalRows":140,"groupCount":141,"groups":142,"others":200},3,1,[143],{"slug":144,"group":145,"sourceId":5,"sourceLabel":6,"table":146,"selfRows":140,"metrics":147,"seqs":158,"entrants":162,"cells":172,"outcomes":194,"locators":195,"hardware":196,"wordings":197,"notes":198},"affan2026semanticmeshing-table-1","affan2026semanticmeshing:Table 1","Table 1",[148,153,156],{"label":149,"unit":150,"statistic":151,"alignment":152},"Acc. (cm), inlier RMSE reconstruction to ground truth","cm","RMSE","SE3",{"label":154,"unit":155,"statistic":84,"alignment":152},"Comp. (%), recall of ground-truth points within tau","%",{"label":157,"unit":155,"statistic":84,"alignment":152},"F1 (%) at tau = 0.30 m",[159],{"dataset":95,"sequence":160,"environment":161},"Christ Church College (sequence 2)","cultural-heritage college buildings (described as indoor in Sec. 5.2)",[163,167,170],{"name":164,"methodId":165,"linkable":166,"proposed":69,"self":69},"ImMesh","lin2023immesh",true,{"name":168,"methodId":169,"linkable":166,"proposed":69,"self":69},"Voxblox","oleynikova2017voxblox",{"name":171,"methodId":5,"linkable":166,"proposed":166,"self":166},"Ours",[173,177,179,182,184,186,188,190,192],[174,174,174,175,176,174,176,176,174],0,17.4,-1,[141,174,174,178,176,174,176,176,174],17.24,[180,174,174,181,176,174,176,176,174],2,14.54,[174,141,174,183,176,174,176,176,174],95.99,[141,141,174,185,176,174,176,176,174],97.17,[180,141,174,187,176,174,176,176,174],98.55,[174,180,174,189,176,174,176,176,174],96.31,[141,180,174,191,176,174,176,176,174],96.98,[180,180,174,193,176,174,176,176,174],98.58,[],[146],[],[],[199],"Meshes sampled to 500,000 points, manually coarse-registered then ICP-aligned to TLS ground truth; outlier radius filter; inlier threshold tau = 0.30 m; accuracy as inlier RMSE, completeness as recall percentage, F1 from precision and recall at tau",[],1790510665791]