Double window optimisation for constant time visual SLAM
The DOI matches its Crossref record; 21 of the 21 references it cites were found in a registry or an index.
Abstract
We present a novel and general optimisation framework for visual SLAM, which scales for both local, highly accurate reconstruction and large-scale motion with long loop closures. We take a two-level approach that combines accurate pose-point constraints in the primary region of interest with a stabilising periphery of pose-pose soft constraints. Our algorithm automatically builds a suitable connected graph of keyposes and constraints, dynamically selects inner and outer window membership and optimises both simultaneously. We demonstrate in extensive simulation experiments that our method approaches the accuracy of offline bundle adjustment while maintaining constant-time operation, even in the hard case of very loopy monocular camera motion. Furthermore, we present a set of real experiments for various types of visual sensor and motion, including large scale SLAM with both monocular and stereo cameras, loopy local browsing with either monocular or RGB-D cameras, and dense RGB-D object model building.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“inspired by [223], Strasdat et al. [235] take a two-stage approach and optimize first a local pose-features graph and then a pose-pose graph; Williams et al. [259] split factor graph optimization in a high-frequency filter and lowfrequency smoother, which are periodically synchronized.”
From Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age (2016) · cites this paper as [235]
Citations
Cites
Machine Learning for High-Speed Corner Detection
Rosten, E., & Drummond, T. (2006) · Lecture Notes in Computer Science · doi:10.1007/11744023_34 · Cited by 4,555
Parallel Tracking and Mapping for Small AR Workspaces
Klein, G., & Murray, D. (2007) · 2007 6th IEEE and ACM International Symposium on Mixed and Augmented Reality · doi:10.1109/ismar.2007.4538852 · Cited by 4,346 · cited 3×
“For example, PTAM [6] runs full bundle adjustment in a background thread, which limits its scale to small workspaces.”
BackgroundClosed-form solution of absolute orientation using unit quaternions
Horn, B. K. P. (1987) · Journal of the Optical Society of America A · doi:10.1364/josaa.4.000629 · Cited by 4,116 · cited 2×
“Given a set of three 3D-3D corresponences, a relative pose Ti,ref is uniquely defined [3].”
BRIEF: Binary Robust Independent Elementary Features
Calonder, M., Lepetit, V., Strecha, C., & Fua, P. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-15561-1_56 · Cited by 3,597 · cited 1×
“Afterwards, we use guided BRIEF [1] matching between measurements and candidate points seeded at FAST [15] corners – followed by 3D to 3D robust pose refinement to produce an initial pose estimate.”
MethodScalable Recognition with a Vocabulary Tree
Nister, D., & Stewenius, H. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06) · doi:10.1109/cvpr.2006.264 · Cited by 3,579 · cited 1×
“Candidates for large loop closures can be efficiently detected using appearance information only [2, 14].”
BackgroundPhoto tourism: exploring photo collections in 3D
Snavely, N., Seitz, S. M., & Szeliski, R. (2006) · ACM Transactions on Graphics · doi:10.1145/1141911.1141964 · Cited by 2,866
G2o: A general framework for graph optimization
Kummerle, R., Grisetti, G., Strasdat, H., Konolige, K., & Burgard, W. (2011) · 2011 IEEE International Conference on Robotics and Automation · doi:10.1109/icra.2011.5979949 · Cited by 1,982 · cited 2×
“Both methods are implemented using the efficient state-of-the-art sparse graph optimizer g(2)o [9] and executed on a single core of an Intel(R) Core(TM) i7 960 desktop computer.”
MethodSelf-citationPhoto tourism: exploring photo collections in 3D
Snavely, N., Seitz, S. M., & Szeliski, R. (2006) · ACM SIGGRAPH 2006 Papers on - SIGGRAPH '06 · doi:10.1145/1179352.1141964 · Cited by 696
FrameSLAM: From Bundle Adjustment to Real-Time Visual Mapping
Konolige, K., & Agrawal, M. (2008) · IEEE Transactions on Robotics · doi:10.1109/tro.2008.2004832 · Cited by 532 · cited 3×
“Instead of estimating this uncertainty accurately using proper marginalisation [8], we suggest to approximate ΛTji coarsely:”
BackgroundMethodSelf-citationReal Time Localization and 3D Reconstruction
Mouragnon, E., Lhuillier, M., Dhome, M., Dekeyser, F., & Sayd, P. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06) · doi:10.1109/cvpr.2006.236 · Cited by 396 · cited 1×
“In general, keyframes are fixed at the boundary of the active window, and all those keyframes with co-visible points are typically included [13].”
BackgroundOn measuring the accuracy of SLAM algorithms
Kümmerle, R., Steder, B., Dornhege, C., Ruhnke, M., Grisetti, G., Stachniss, C., & Kleiner, A. (2009) · Autonomous Robots · doi:10.1007/s10514-009-9155-6 · Cited by 331 · cited 1×
“[10] and define a relative error in terms of the relative differences ∆Tij := TiT j between two absolute poses.”
BackgroundScale Drift-Aware Large Scale Monocular SLAM
Strasdat, H., M. M. Montiel, J., & Davison, A. (2010) · Robotics: Science and Systems VI · doi:10.15607/rss.2010.vi.010 · Cited by 323 · cited 5×
“In order to deal with scale drift, we represent the camera poses T for monocular SLAM using the Lie group of similarity transformations Sim(3) [18]:”
BackgroundMethodInfluentialSelf-citationScale Drift-Aware Large Scale Monocular SLAM
Strasdat, H., Montiel, J. M. M., & Davison, A. J. (2011) · Robotics · doi:10.7551/mitpress/9123.003.0014 · Cited by 320
Self-citationHighly scalable appearance-only SLAM - FAB-MAP 2.0
Cummins, M., & Newman, P. (2009) · Robotics: Science and Systems V · doi:10.15607/rss.2009.v.039 · Cited by 282 · cited 1×
“Candidates for large loop closures can be efficiently detected using appearance information only [2, 14].”
BackgroundVast-scale Outdoor Navigation Using Adaptive Relative Bundle Adjustment
Sibley, G., Mei, C., Reid, I., & Newman, P. (2010) · The International Journal of Robotics Research · doi:10.1177/0278364910369268 · Cited by 156
Sparse Sparse Bundle Adjustment
Konolige, K. (2010) · British Machine Vision Conference (BMVC) · doi:10.5244/c.24.102 · Cited by 151 · cited 1×
“Full bundle adjustment in VSLAM, while improving rapidly in absolute computational time [5, 7], still suffers from linear to cubic time in the number of variables (depending on particulars of the system), thus limiting its use in large-scale operation.”
BackgroundSelf-citationAdaptive relative bundle adjustment
Sibley, G., Mei, C., Reid, I., & Newman, P. (2009) · Robotics: Science and Systems V · doi:10.15607/rss.2009.v.023 · Cited by 139 · cited 9×
“To work in constant time, RBA makes the active window assumption.”
BackgroundInfluentialMulti-robot mapping using manifold representations
Howard, A. (2004) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/robot.2004.1308933 · Cited by 110 · cited 1×
“One proposal for how maps could work in both cases is to consider them as manifolds; that is, locally Euclidean but globally topological [4].”
BackgroundMei, C., Sibley, G., & Newman, P. (2010) · 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems · doi:10.1109/iros.2010.5652266 · Cited by 67 · cited 1×
“Posepose constraints are defined by covisiblity [12].”
Lim, J., Frahm, J.-M., & Pollefeys, M. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995511 · Cited by 53 · cited 4×
“Instead of anchoring the 3D points with respect to a keyframe [16, 11], we express them simply in a global coordinate frame, and thus can avoid a point management overhead.”
BackgroundMethodInfluentialPushing the envelope of modern methods for bundle adjustment
Jeong, Y., Nistér, D., Steedly, D., Szeliski, R., & Kweon, I.-S. (2010) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2010.5539795 · Cited by 48 · cited 2×
“Instead, we let the damping factor of Levenberg-Marquardt takes care of the gauge freedom [5].”
BackgroundMethod
Research works printed but not linked: 1
- rosten, 2006, Machine learning for highspeed corner detection, Proceedings of the European Conference on Computer Vision (ECCV)
Cited by
ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras
Mur-Artal, R., & Tardos, J. D. (2017) · IEEE Transactions on Robotics · doi:10.1109/tro.2017.2705103 · Cited by 6,898
“[8] performs a joint optimization of BA (point-pose constraints) in an inner window of keyframes and pose graph (pose-pose constraints) in an outer window.”
BackgroundMethodInfluentialORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · IEEE Transactions on Robotics · doi:10.1109/tro.2015.2463671 · Cited by 6,778
“[6], [7], the optimization framework g2o by Kuemmerle et al.”
BackgroundMethodInfluentialSelf-citationORB-SLAM3: An Accurate Open-Source Library for Visual, Visual–Inertial, and Multimap SLAM
Campos, C., Elvira, R., Rodriguez, J. J. G., M. Montiel, J. M., & D. Tardos, J. (2021) · IEEE Transactions on Robotics · doi:10.1109/tro.2021.3075644 · Cited by 4,274
Self-citationPast, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age
Cadena, C., Carlone, L., Carrillo, H., Latif, Y., Scaramuzza, D., Neira, J., Reid, I., & Leonard, J. J. (2016) · IEEE Transactions on Robotics · doi:10.1109/tro.2016.2624754 · Cited by 3,596
“[235] take a two-stage approach and optimize first a local pose-features graph and then a pose-pose graph; Williams et al.”
ORB-SLAM: A Versatile and Accurate Monocular SLAM System
Mur-Artal, R., Montiel, J. M. M., & Tardos, J. D. (2015) · Zaguan (Universidad de Zaragoza) · Cited by 3,153
Self-citation
Metadata
- Authors
- Strasdat, H., Davison, A. J., Montiel, J. M. M., & Konolige, K.
- Affiliations
- Imperial College London (GB)
- Universidad de Zaragoza (ES)
- Venue
- 2011 International Conference on Computer Vision
- Publisher
- IEEE
- Published
- 2011-11
- Pages
- 2352-2359
- OpenAlex
- W2154280780
- Cited by
- 296
- References
- 21
- Citations by year
- 2026: 2 · 2025: 5 · 2024: 9 · 2023: 13 · 2022: 25 · 2021: 20 · 2020: 30 · 2019: 33 · 2018: 31 · 2017: 30 · 2016: 23 · 2015: 34 · 2014: 15 · 2013: 21 · 2012: 5
- FWCI
- 25.35
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Earth and Planetary Sciences › Geology › 3D Surveying and Cultural Heritage
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Keywords
- visual SLAM, loop closure, monocular camera, stereo camera, RGB-D, dense 3D reconstruction, bundle adjustment
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.