[{"data":1,"prerenderedAt":112},["ShallowReactive",2],{"method-dmvio2022":3},{"method":4,"reference":63,"equipment":82,"figures":111,"results":87},{"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":22,"limitations":27,"sensors":32,"platform":35,"estimator":39,"association":40,"timeModel":41,"deskew":42,"loopClosure":43,"globalOptimization":44,"mapRepresentation":45,"prior":46,"outputGeometry":47,"compute":48,"codeUrl":49,"codeLicense":50,"relatedVersions":51},"dmvio2022","von Stumberg & Cremers, 2022","DM-VIO","DM-VIO: Delayed Marginalization Visual-Inertial Odometry",2022,"recent","C08","odometry_with_local_mapping","DM-VIO 是單目視覺慣性里程計，以 DSO 的直接光度光束法平差為核心，加入 IMU 預積分並把尺度與重力方向作為顯式變數持續最佳化。作者提出延遲邊際化：另外維護一個延遲 100 個關鍵影格才邊際化的因子圖，可在其中加入 IMU 因子做位姿圖光束法平差（PGBA），以完整的光度不確定度初始化 IMU，並重新推進該圖得到含 IMU 資訊的邊際化先驗；尺度大幅改變時也能替換邊際化先驗。另以動態光度權重在影像品質差時提高 IMU 比重。","DM-VIO is a direct monocular VIO that keeps a delayed marginalisation graph to run pose graph bundle adjustment for IMU initialisation with the full photometric uncertainty, to readvance an IMU-informed marginalisation prior, and to replace it when scale changes, while scale and gravity remain optimised in the main system.","full_text_reviewed","peer_reviewed_published","supplementary","論文未在營建工地測試；評估涵蓋 EuRoC（無人機）、TUM-VI（大範圍室內外手持資料，含沿管道滑下的序列）與 4Seasons（車輛）。單目相機加 IMU 即可得到公制尺度且漂移低的軌跡，對以手持或頭戴裝置在施工中建物內巡檢定位有參考價值；但只輸出稀疏點雲且無迴圈閉合，長距離戶外仍有數公尺至數十公尺漂移（推論）。",[20,21],"public_benchmark","independent_reference",[23,24,25,26],"Lowest average EuRoC RMSE among the compared VIO systems (0.069 m, versus 0.072 m for stereo Basalt and 0.089 m for VI-DSO) with average scale error 0.6% (Table I)","On TUM-VI best on 16 sequences with mean normalised drift 0.472% versus 0.939% for Basalt (Table II)","Outperforms stereo-inertial ORB-SLAM3 and Basalt on 4Seasons despite monocular input and no loop closure (Sec. IV-C; Fig. 5, plotted)","Delayed marginalisation costs only about 0.44 ms (0.8%) in the keyframe thread (Sec. IV-A)",[28,29,30,31],"Odometry only: ORB-SLAM3 with loop closure is more accurate on some TUM-VI sequences (Sec. IV-B; Fig. 4)","Scale is not observable during constant-velocity motion, so monocular VIO initialisation remains difficult in automotive scenes (Secs. I, IV-C)","Large drift remains on long TUM-VI outdoor sequences (for example 123.24 m on outdoors1, 2656 m long) (Table II)","4Seasons required cropping the car hood and modifying the visual initialiser (zero prior on x-y translation, keyframe threshold) (Sec. IV-C)",[33,34],"monocular camera","IMU",[36,37,38],"UAV (EuRoC MAV)","handheld (TUM-VI, large-scale indoor and outdoor)","vehicle (4Seasons)","direct (DSO-based) photometric visual-inertial bundle adjustment over up to 8 active keyframes with IMU preintegration, dynamic photometric weight, and explicit scale and gravity-direction variables; Schur-complement partial marginalisation with FEJ plus a second delayed marginalisation graph (delay 100) used for pose graph bundle adjustment (PGBA) IMU initialisation and marginalisation replacement; Levenberg-Marquardt with SIMD photometric code and GTSAM (Sec. III)","direct: photometric residuals of sparse DSO points over a neighbourhood pattern with affine brightness and exposure; no feature matching (Sec. III-B)","discrete keyframes with IMU preintegration between keyframes (Sec. III-B)","not_applicable","none (odometry; loop closure and map reuse named as future work)","none","sparse point cloud of inverse-depth points hosted in active keyframes (Fig. 1 point clouds)","priors on first pose and gravity direction; IMU noise parameters from calibration or data sheets (Secs. III-B, IV-C)","metric-scale camera and IMU trajectory with sparse point cloud (Fig. 1)","real-time mode on a MacBook Pro 2013 (i7 at 2.3 GHz) without GPU: tracking 10.34 ms per frame and keyframe processing 53.67 ms on average; delayed marginalisation overhead 0.44 ms (0.8%) in the keyframe thread (Sec. IV; IV-A)","https:\u002F\u002Fgithub.com\u002Flukasvst\u002Fdm-vio","GPL-3.0 (LICENSE file)",[52,56,59],{"relation":53,"title":54,"doi_or_url":55},"preprint","DM-VIO: Delayed Marginalization Visual-Inertial Odometry (arXiv v1, accepted RA-L version)","https:\u002F\u002Farxiv.org\u002Fabs\u002F2201.04114",{"relation":57,"title":58,"doi_or_url":49},"code_release","lukasvst\u002Fdm-vio (project page vision.in.tum.de\u002Fdm-vio; supplementary with ablations and runtime)",{"relation":60,"title":61,"doi_or_url":62},"predecessor_method","VI-DSO: Direct sparse visual-inertial odometry using dynamic marginalization (ICRA 2018), cited as [6] and compared in Table I","10.1109\u002FICRA.2018.8462905",{"id":5,"kind":64,"shortName":7,"title":8,"authors":65,"year":9,"venue":68,"venueType":69,"publisher":70,"volumeIssuePages":71,"doi":72,"arxivId":73,"url":55,"firstPublicDate":74,"publicationStatus":16,"metadataStatus":75,"fulltextStatus":15,"era":10,"classicReason":42,"codeUrl":49,"cluster":11,"topics":76,"mdpi":77,"verification":78,"label":6,"fulltextRoute":79,"versionRead":80,"addedByCensus":81},"method",[66,67],"Lukas von Stumberg","Daniel Cremers","IEEE Robotics and Automation Letters","journal","IEEE","7(2):1408-1415","10.1109\u002Flra.2021.3140129","2201.04114","2022-01-04","metadata_verified",[11],false,"corrected","arXiv","arXiv v1 (2022-01-11), accepted RA-L version with IEEE copyright notice; IEEE version of record not read",true,[83,90,93,100,106],{"category":84,"model":85,"canonical":85,"role":86,"dataset":87,"specs":88,"locator":89},"compute","MacBook Pro 2013 (i7 at 2.3GHz)","compute for runtime",null,"real-time mode without GPU; same machine as in the VI-DSO paper","Sec. IV",{"category":84,"model":91,"canonical":91,"role":86,"dataset":87,"specs":92,"locator":89},"Intel Core i7-7700K at 4.2GHz desktop","used only to run ORB-SLAM3 (not supported on macOS)",{"category":94,"model":95,"canonical":95,"role":96,"dataset":97,"specs":98,"locator":99},"mobile_scanner_device","TUM-VI handheld visual-inertial device (model not reported in this paper)","dataset sensor","TUM-VI","large-scale indoor and outdoor handheld sequences, including sliding down a tube","Sec. IV-B",{"category":101,"model":102,"canonical":102,"role":96,"dataset":103,"specs":104,"locator":105},"camera","4Seasons visual-inertial sensor (model not reported)","4Seasons","well time-synchronised; bottom 96 pixels cropped because of the car hood; IMU noise read from the data-sheet Allan variance plot","Sec. IV-C",{"category":101,"model":107,"canonical":107,"role":96,"dataset":108,"specs":109,"locator":110},"EuRoC MAV camera and IMU (models not reported in this paper)","EuRoC MAV","monocular images and IMU data recorded by a flying drone","Secs. I, IV-A",[],1790510657969]