[{"data":1,"prerenderedAt":436},["ShallowReactive",2],{"method-tuna2024xicp":3},{"method":4,"reference":60,"equipment":83,"figures":126,"results":127},{"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":24,"limitations":30,"sensors":38,"platform":42,"estimator":44,"association":45,"timeModel":46,"deskew":47,"loopClosure":48,"globalOptimization":48,"mapRepresentation":49,"prior":50,"outputGeometry":51,"compute":52,"codeUrl":53,"codeLicense":54,"relatedVersions":55},"tuna2024xicp","Tuna et al., 2024","X-ICP","X-ICP: Localizability-Aware LiDAR Registration for Robust Localization in Extreme Environments",2024,"recent","C02","registration_component","X-ICP 針對 LiDAR 在幾何資訊不足環境（隧道、開放平面、狹窄走廊）中 ICP 沿弱約束方向發散的問題，先利用掃描與地圖的對應，分析各最佳化主方向的對齊強度，細緻判定可定位性（localizability）。再將此分析整合進掃描對地圖的點對平面 ICP，以約束最佳化控制或凍結退化方向的位姿更新。作者以 ANYmal 足式機器人在地下礦坑、營建工地與城市公園實測，並以 Leica RTC360 地面掃描作為礦坑與公園的參考地圖。","X-ICP detects fine-grained LiDAR localizability from scan-to-map correspondences and constrains point-to-plane ICP updates along degenerate directions; tested on a legged robot in a mine, a construction site and a park.","full_text_reviewed","peer_reviewed_published","main_body","作者實測：瑞士 Rumlang 大型營建工地 153 m 路線（開放平面、原地旋轉），以地圖一致性定性比較；Seemuhle 地下礦坑隧道段以 TLS 參考地圖量化。屬單一研究、單一工地，不可推廣為一般工地性能。",[20,21,22,23],"simulation","real_construction_site","underground_or_tunnel","independent_reference",[25,26,27,28,29],"accurate localizability detection and robust pose estimation without environment-specific parameter tuning, with the same kappa_1 = 250, kappa_2 = 180, kappa_3 = 35 for all environments and sensors (abstract, Sec. V-C)","Seemuhle VLP-16 run: APE translation 2.05 (1.23) m with first-15 m alignment vs 3.36 (1.74) m for Zhang et al. and 5.79 (5.26) m for Hinduja et al., last-position error 0.27 m vs 6.37 m and 24.17 m (Table I)","RPE per 10 m 0.17 m vs 0.20 m and 0.26 m (Table II)","partial localizability improves translation over the binary variant Xs-ICP, last-position error 0.27 m vs 5.34 m (Tables III to IV)","consistent map at the Rumlang construction site where the eigenvalue-threshold baseline performed poorly (Sec. VII-E)",[31,32,33,34,35,36,37],"sensitive to initial-guess quality like related methods, and with a truly bad prior the registration cannot be solved reliably (Sec. VI-B, VIII)","the filtering parameter kappa_f must be re-adjusted per LiDAR (cos 80 deg for VLP-16, cos 60 deg for OS0-128) (Sec. V-B), and the authors plan to improve the sensor-dependent selection of kappa_1 with point-wise confidence weights (Sec. VIII)","no ground-truth map or quantitative error is reported for the Rumlang construction-site run, which is judged from map appearance (Sec. VII-E)","baselines were re-implemented by the authors, and the Zhang et al. eigenvalue threshold was 120 in all experiments except Opfikon, where it was tuned to 200 (Sec. VII-B, VII-F)","even X-ICP shows about 2 m mean APE over the 521.8 m mine run (Table I","reviewer observation)","code release not verified (project website only)",[39,40,41],"3D LiDAR (Velodyne VLP-16; Ouster OS0-128)","IMU","leg joint encoders (leg odometry prior)",[43,20],"legged","scan-to-map point-to-plane ICP with localizability-driven constrained optimization (Lagrange multipliers) that fixes or limits updates along degenerate directions (abstract, Sec. VI)","scan-to-map correspondences analysed against principal optimization directions for fine-grained localizability (abstract)","discrete poses","point cloud motion compensation done at the LiDAR driver level with the leg-odometry pose estimates in the transformation tree; the pose prior is used to transform and undistort the input cloud","none","point cloud map in a libpointmatcher-based registration framework (Sec. VII-A)","leg odometry prior (IMU + joint encoders) as initial guess (Sec. VII-A)","pose updates; resulting point cloud maps compared with TLS reference maps (Sec. VII)","mapping pipeline runs at 5 Hz (Sec. IV); single-threaded scan-to-map registration on the Rumlang data takes 32.19 (10.7) ms with X-ICP on an Intel i7-9750H laptop, stated to be equivalent to the robot computer, and 12.65 (3.51) ms on an Intel i9-13900K, vs 20.05 (2.21) ms and 11.1 (1.45) ms for the baseline without localizability awareness, whose statistics cover only the first 150 s of the run (Table V); localizability overhead per ICP iteration shown only as a plot (Fig. 20)",null,"not_verified",[56],{"relation":57,"title":58,"doi_or_url":59},"preprint","arXiv:2211.16335 (v1 2022-11-29; v4 2024-02-18)","https:\u002F\u002Farxiv.org\u002Fabs\u002F2211.16335",{"id":5,"kind":61,"shortName":7,"title":8,"authors":62,"year":9,"venue":68,"venueType":69,"publisher":70,"volumeIssuePages":71,"doi":72,"arxivId":73,"url":59,"firstPublicDate":74,"publicationStatus":16,"metadataStatus":75,"fulltextStatus":15,"era":10,"classicReason":76,"codeUrl":53,"cluster":11,"topics":77,"mdpi":79,"verification":80,"label":6,"fulltextRoute":81,"versionRead":82,"addedByCensus":79},"method",[63,64,65,66,67],"Turcan Tuna","Julian Nubert","Yoshua Nava","Shehryar Khattak","Marco Hutter","IEEE Transactions on Robotics","journal","IEEE","40:452-471","10.1109\u002Ftro.2023.3335691","2211.16335","2022-11-29","metadata_verified","not_applicable",[11,78],"C13",false,"confirmed","arXiv","arXiv 2211.16335v4 (18 Feb 2024), accepted version of IEEE T-RO vol. 40, 2024, 20 pages; T-RO version of record not compared",[84,90,95,100,104,109,116,122],{"category":85,"model":86,"canonical":86,"role":87,"dataset":53,"specs":88,"locator":89},"lidar","Velodyne VLP-16","method input","16-beam (sparse) LiDAR on ANYmal-C; used in all field experiments","Sec. VII-A, VII-G3",{"category":85,"model":91,"canonical":91,"role":87,"dataset":92,"specs":93,"locator":94},"Ouster OS0-128","Seemuhle mine repeat run","much higher point density and larger FoV than VLP-16; 128-beam (dense) LiDAR","Sec. VII-D2, VII-G3",{"category":96,"model":97,"canonical":97,"role":87,"dataset":53,"specs":98,"locator":99},"imu","inertial measurement unit (IMU)","not_reported","Sec. VII-A",{"category":101,"model":102,"canonical":102,"role":87,"dataset":53,"specs":103,"locator":99},"wheel_or_leg_odometry","joint encoders","used with the IMU by ANYmal's leg odometry module (ref. [69]) to provide the registration prior",{"category":105,"model":106,"canonical":106,"role":87,"dataset":53,"specs":107,"locator":108},"platform","ANYmal-C","legged robot","Sec. VII-A, Fig. 6",{"category":110,"model":111,"canonical":112,"role":113,"dataset":114,"specs":98,"locator":115},"tls_scanner","Leica RTC 360","Leica RTC360","reference or ground truth","Seemuhle mine and Opfikon City Park ground-truth maps","Sec. VII-D, VII-F, Fig. 6-c",{"category":117,"model":118,"canonical":118,"role":119,"dataset":53,"specs":120,"locator":121},"compute","Intel i7-9750H","compute for runtime","laptop CPU equivalent to the one on the robot; single-threaded timing","Sec. VII-A, Table V",{"category":117,"model":123,"canonical":123,"role":119,"dataset":53,"specs":124,"locator":125},"Intel i9-13900K","desktop-class CPU; single-threaded timing","Sec. VII-G3, Table V",[],{"totalRows":128,"groupCount":129,"groups":130,"others":388},65,12,[131,223,303,344],{"slug":132,"group":133,"sourceId":5,"sourceLabel":6,"table":134,"selfRows":135,"metrics":136,"seqs":154,"entrants":163,"cells":169,"outcomes":216,"locators":217,"hardware":218,"wordings":219,"notes":220},"tuna2024xicp-table-iii","tuna2024xicp:Table III","Table III",18,[137,141,143,146,147,149,150,151,152],{"label":138,"unit":139,"statistic":140,"alignment":98},"APE Translation mu(sigma) [m]","m","mean",{"label":138,"unit":139,"statistic":142,"alignment":98},"std",{"label":144,"unit":145,"statistic":140,"alignment":98},"APE Rotation mu(sigma) [deg]","deg",{"label":144,"unit":145,"statistic":142,"alignment":98},{"label":138,"unit":139,"statistic":140,"alignment":148},"first-pose",{"label":138,"unit":139,"statistic":142,"alignment":148},{"label":144,"unit":145,"statistic":140,"alignment":148},{"label":144,"unit":145,"statistic":142,"alignment":148},{"label":153,"unit":139,"statistic":98,"alignment":98},"Last Position Error [m]",[155,159,161],{"dataset":156,"sequence":157,"environment":158},"Seemuhle underground mine (authors' data)","VLP-16 run; first 15 m alignment","underground mine tunnel",{"dataset":156,"sequence":160,"environment":158},"VLP-16 run; origin alignment",{"dataset":156,"sequence":162,"environment":158},"VLP-16 run; full 521.8 m traverse",[164,167],{"name":165,"methodId":5,"linkable":166,"proposed":166,"self":166},"X-ICP (Proposed)",true,{"name":168,"methodId":5,"linkable":166,"proposed":166,"self":166},"Xs-ICP (Proposed)",[170,174,177,180,183,186,189,192,195,198,200,202,204,206,208,210,212,214],[171,171,171,172,173,171,173,173,171],0,2.05,-1,[171,175,171,176,173,171,173,173,175],1,1.23,[171,178,171,179,173,171,173,173,175],2,2.55,[171,181,171,182,173,171,173,173,175],3,0.76,[171,184,175,185,173,171,173,173,175],4,2.45,[171,187,175,188,173,171,173,173,175],5,1.35,[171,190,175,191,173,171,173,173,175],6,2.5,[171,193,175,194,173,171,173,173,175],7,1.03,[171,196,178,197,173,171,173,173,175],8,0.27,[175,171,171,199,173,171,173,173,175],2.29,[175,175,171,201,173,171,173,173,175],1.22,[175,178,171,203,173,171,173,173,175],3.06,[175,181,171,205,173,171,173,173,175],1.19,[175,184,175,207,173,171,173,173,175],2.68,[175,187,175,209,173,171,173,173,175],1.26,[175,190,175,211,173,171,173,173,175],3.08,[175,193,175,213,173,171,173,173,175],1.3,[175,196,178,215,173,171,173,173,175],5.34,[],[134],[],[],[221,222],"Ablation, Seemuhle VLP-16 APE (same protocol as Table I); Xs-ICP drops the partial-localizability category and categorizes only in the first ICP iteration","Same setting as other Table III rows",{"slug":224,"group":225,"sourceId":5,"sourceLabel":6,"table":226,"selfRows":227,"metrics":228,"seqs":238,"entrants":242,"cells":250,"outcomes":296,"locators":297,"hardware":298,"wordings":299,"notes":300},"tuna2024xicp-table-i","tuna2024xicp:Table I","Table I",9,[229,230,231,232,233,234,235,236,237],{"label":138,"unit":139,"statistic":140,"alignment":98},{"label":138,"unit":139,"statistic":142,"alignment":98},{"label":144,"unit":145,"statistic":140,"alignment":98},{"label":144,"unit":145,"statistic":142,"alignment":98},{"label":138,"unit":139,"statistic":140,"alignment":148},{"label":138,"unit":139,"statistic":142,"alignment":148},{"label":144,"unit":145,"statistic":140,"alignment":148},{"label":144,"unit":145,"statistic":142,"alignment":148},{"label":153,"unit":139,"statistic":98,"alignment":98},[239,240,241],{"dataset":156,"sequence":157,"environment":158},{"dataset":156,"sequence":160,"environment":158},{"dataset":156,"sequence":162,"environment":158},[243,244,247],{"name":165,"methodId":5,"linkable":166,"proposed":166,"self":166},{"name":245,"methodId":246,"linkable":166,"proposed":79,"self":79},"Zhang et al. [12]","zhang2016degeneracy",{"name":248,"methodId":249,"linkable":166,"proposed":79,"self":79},"Hinduja et al. [17]","hinduja2019degeneracy",[251,252,253,254,255,256,257,258,259,260,262,264,266,268,270,272,274,276,278,280,282,284,286,288,290,292,294],[171,171,171,172,173,171,173,173,171],[171,175,171,176,173,171,173,173,175],[171,178,171,179,173,171,173,173,175],[171,181,171,182,173,171,173,173,175],[171,184,175,185,173,171,173,173,175],[171,187,175,188,173,171,173,173,175],[171,190,175,191,173,171,173,173,175],[171,193,175,194,173,171,173,173,175],[171,196,178,197,173,171,173,173,175],[175,171,171,261,173,171,173,173,175],3.36,[175,175,171,263,173,171,173,173,175],1.74,[175,178,171,265,173,171,173,173,175],4.06,[175,181,171,267,173,171,173,173,175],1.37,[175,184,175,269,173,171,173,173,175],3.73,[175,187,175,271,173,171,173,173,175],1.8,[175,190,175,273,173,171,173,173,175],4.11,[175,193,175,275,173,171,173,173,175],1.52,[175,196,178,277,173,171,173,173,175],6.37,[178,171,171,279,173,171,173,173,175],5.79,[178,175,171,281,173,171,173,173,175],5.26,[178,178,171,283,173,171,173,173,175],7.67,[178,181,171,285,173,171,173,173,175],4.72,[178,184,175,287,173,171,173,173,175],8.16,[178,187,175,289,173,171,173,173,175],4.83,[178,190,175,291,173,171,173,173,175],8.03,[178,193,175,293,173,171,173,173,175],4.73,[178,196,178,295,173,171,173,173,175],24.17,[],[226],[],[],[301,302],"Seemuhle underground mine, ANYmal with VLP-16, 521.8 m; APE via EVO against Leica RTC 360 ground truth, mu (sigma); 'first 15 m' = trajectory aligned on the first 15 m (about 200 poses), 'origin' = aligned at the first pose; plus last-position error","Same setting as other Table I rows",{"slug":304,"group":305,"sourceId":5,"sourceLabel":6,"table":306,"selfRows":196,"metrics":307,"seqs":314,"entrants":317,"cells":320,"outcomes":337,"locators":338,"hardware":339,"wordings":340,"notes":341},"tuna2024xicp-table-iv","tuna2024xicp:Table IV","Table IV",[308,310,311,313],{"label":309,"unit":139,"statistic":140,"alignment":48},"RPE Translation mu(sigma) [m] per 10 m",{"label":309,"unit":139,"statistic":142,"alignment":48},{"label":312,"unit":145,"statistic":140,"alignment":48},"RPE Rotation mu(sigma) [deg] per 10 m",{"label":312,"unit":145,"statistic":142,"alignment":48},[315],{"dataset":156,"sequence":316,"environment":158},"VLP-16 run",[318,319],{"name":165,"methodId":5,"linkable":166,"proposed":166,"self":166},{"name":168,"methodId":5,"linkable":166,"proposed":166,"self":166},[321,323,325,327,329,331,333,335],[171,171,171,322,173,171,173,173,171],0.17,[171,175,171,324,173,171,173,173,175],0.12,[171,178,171,326,173,171,173,173,175],0.86,[171,181,171,328,173,171,173,173,175],0.42,[175,171,171,330,173,171,173,173,175],0.19,[175,175,171,332,173,171,173,173,175],0.13,[175,178,171,334,173,171,173,173,175],0.85,[175,181,171,336,173,171,173,173,175],0.47,[],[306],[],[],[342,343],"Ablation, RPE per 10 m traversed distance, Seemuhle VLP-16, mu (sigma)","Same setting as other Table IV rows",{"slug":345,"group":346,"sourceId":5,"sourceLabel":6,"table":347,"selfRows":196,"metrics":348,"seqs":353,"entrants":358,"cells":362,"outcomes":379,"locators":380,"hardware":381,"wordings":384,"notes":385},"tuna2024xicp-table-v","tuna2024xicp:Table V","Table V",[349,352],{"label":350,"unit":351,"statistic":140,"alignment":48},"scan-to-map registration time mu (sigma) [ms]","ms",{"label":350,"unit":351,"statistic":142,"alignment":48},[354],{"dataset":355,"sequence":356,"environment":357},"Rumlang construction site (authors' data)","153 m run","construction site, open planar area",[359,360],{"name":7,"methodId":5,"linkable":166,"proposed":166,"self":166},{"name":361,"methodId":5,"linkable":166,"proposed":166,"self":166},"Xs-ICP",[363,365,367,369,371,373,375,377],[171,171,171,364,173,171,171,173,171],12.65,[171,175,171,366,173,171,171,173,175],3.51,[171,171,171,368,173,171,175,173,175],32.19,[171,175,171,370,173,171,175,173,175],10.7,[175,171,171,372,173,171,171,173,175],11.14,[175,175,171,374,173,171,171,173,175],3.1,[175,171,171,376,173,171,175,173,175],29.42,[175,175,171,378,173,171,175,173,175],9.35,[],[347],[382,383],"Intel i9-13900K, single thread","Intel i7-9750H, single thread",[],[386,387],"Scan-to-map registration time per scan on the Rumlang construction-site data, single-threaded, mu (sigma); baseline statistics computed only until 150 s into the run to avoid the degenerate part","Same setting as other Table V rows",[389,395,400,407,413,419,425,430],{"group":390,"slug":391,"sourceLabel":392,"table":347,"selfRows":184,"datasets":393},"genzicp2025:Table V","genzicp2025-table-v","Lee et al., 2025a",[394],"Ground-Challenge",{"group":396,"slug":397,"sourceLabel":6,"table":398,"selfRows":184,"datasets":399},"tuna2024xicp:Table II","tuna2024xicp-table-ii","Table II",[156],{"group":401,"slug":402,"sourceLabel":403,"table":404,"selfRows":184,"datasets":405},"tuna2025informed:Table 3","tuna2025informed-table-3","Tuna et al., 2025","Table 3",[406],"ANYmal simulation",{"group":408,"slug":409,"sourceLabel":403,"table":410,"selfRows":181,"datasets":411},"tuna2025informed:Table 4","tuna2025informed-table-4","Table 4",[412],"ANYmal forest experiment",{"group":414,"slug":415,"sourceLabel":403,"table":416,"selfRows":181,"datasets":417},"tuna2025informed:Table 5","tuna2025informed-table-5","Table 5",[418],"ENWIDE (Ulmberg bicycle tunnel)",{"group":420,"slug":421,"sourceLabel":392,"table":422,"selfRows":178,"datasets":423},"genzicp2025:Table VI","genzicp2025-table-vi","Table VI",[424],"SubT-MRS",{"group":426,"slug":427,"sourceLabel":392,"table":306,"selfRows":175,"datasets":428},"genzicp2025:Table IV","genzicp2025-table-iv",[429],"HILTI-Oxford 2022",{"group":431,"slug":432,"sourceLabel":403,"table":433,"selfRows":175,"datasets":434},"tuna2025informed:Table 6","tuna2025informed-table-6","Table 6",[435],"HEAP excavator experiment",1790510662179]