{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,6]],"date-time":"2026-06-06T17:10:32Z","timestamp":1780765832784,"version":"3.54.1"},"reference-count":75,"publisher":"SAGE Publications","issue":"13","license":[{"start":{"date-parts":[[2020,7,29]],"date-time":"2020-07-29T00:00:00Z","timestamp":1595980800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"funder":[{"DOI":"10.13039\/501100000038","name":"natural sciences and engineering research council of canada","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100000038","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["The International Journal of Robotics Research"],"published-print":{"date-parts":[[2020,11]]},"abstract":"<jats:p>We present a Gaussian variational inference (GVI) technique that can be applied to large-scale nonlinear batch state estimation problems. The main contribution is to show how to fit both the mean and (inverse) covariance of a Gaussian to the posterior efficiently, by exploiting factorization of the joint likelihood of the state and data, as is common in practical problems. This is different than maximum a posteriori (MAP) estimation, which seeks the point estimate for the state that maximizes the posterior (i.e., the mode). The proposed exactly sparse Gaussian variational inference (ESGVI) technique stores the inverse covariance matrix, which is typically very sparse (e.g., block-tridiagonal for classic state estimation). We show that the only blocks of the (dense) covariance matrix that are required during the calculations correspond to the non-zero blocks of the inverse covariance matrix, and further show how to calculate these blocks efficiently in the general GVI problem. ESGVI operates iteratively, and while we can use analytical derivatives at each iteration, Gaussian cubature can be substituted, thereby producing an efficient derivative-free batch formulation. ESGVI simplifies to precisely the Rauch\u2013Tung\u2013Striebel (RTS) smoother in the batch linear estimation case, but goes beyond the \u2018extended\u2019 RTS smoother in the nonlinear case because it finds the best-fit Gaussian (mean and covariance), not the MAP point estimate. We demonstrate the technique on controlled simulation problems and a batch nonlinear simultaneous localization and mapping problem with an experimental dataset.<\/jats:p>","DOI":"10.1177\/0278364920937608","type":"journal-article","created":{"date-parts":[[2020,7,29]],"date-time":"2020-07-29T04:35:15Z","timestamp":1595997315000},"page":"1473-1502","update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":28,"title":["Exactly sparse Gaussian variational inference with application to derivative-free batch nonlinear state estimation"],"prefix":"10.1177","volume":"39","author":[{"given":"Timothy D","family":"Barfoot","sequence":"first","affiliation":[{"name":"Institute for Aerospace Studies, University of Toronto, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"James R","family":"Forbes","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, McGill University, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David J","family":"Yoon","sequence":"additional","affiliation":[{"name":"Institute for Aerospace Studies, University of Toronto, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2020,7,29]]},"reference":[{"key":"bibr1-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/PL00000559"},{"key":"bibr2-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1016\/j.sigpro.2014.12.013"},{"key":"bibr3-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1162\/089976698300017746"},{"key":"bibr4-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/s10514-015-9455-y"},{"key":"bibr5-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2009.2019800"},{"key":"bibr6-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1017\/9781316671528"},{"key":"bibr7-0278364920937608","unstructured":"Barfoot TD (2020) Multivariate Gaussian variational inference by natural gradient descent. Technical report, Autonomous Space Robotics Lab, University of Toronto."},{"key":"bibr8-0278364920937608","doi-asserted-by":"publisher","DOI":"10.15607\/RSS.2014.X.001"},{"key":"bibr9-0278364920937608","author":"Bayes T","year":"1764","journal-title":"Philosophical Transactions of the Royal Society of London"},{"key":"bibr10-0278364920937608","volume-title":"Pattern Recognition and Machine Learning","author":"Bishop CM","year":"2006"},{"key":"bibr11-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611971484"},{"key":"bibr12-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46484-8_8"},{"key":"bibr13-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TPAS.1978.354538"},{"key":"bibr14-0278364920937608","unstructured":"Brown DC (1958) A solution to the general problem of multiple station analytical stereotriangulation. RCA-MTP Data Reduction Technical Report No. 43 (or AFMTC TR 58-8), Patrick Airforce Base, Florida."},{"key":"bibr15-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1017\/S0962492900002701"},{"key":"bibr16-0278364920937608","author":"Davison AJ","year":"2019","journal-title":"arXiv Preprint"},{"key":"bibr17-0278364920937608","first-page":"4","volume":"12","author":"Dong J","year":"2016","journal-title":"Robotics: Science and Systems"},{"key":"bibr18-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/MRA.2006.1638022"},{"key":"bibr19-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1145\/360680.360704"},{"key":"bibr20-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1145\/358669.358692"},{"issue":"594","key":"bibr21-0278364920937608","first-page":"309","volume":"222","author":"Fisher RA","year":"1922","journal-title":"Philosophical Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character"},{"key":"bibr22-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2015.2454485"},{"key":"bibr23-0278364920937608","doi-asserted-by":"publisher","DOI":"10.3182\/20110828-6-IT-1002.03024"},{"key":"bibr24-0278364920937608","unstructured":"Gauss C (1809) Theoria motus corporum coelestium, perthes et besser, hamburgi 7."},{"key":"bibr25-0278364920937608","first-page":"1","author":"Gauss C","year":"1821","journal-title":"Werke, IV, Koniglichen Gesellschaft der Wissenschaften zu Gottingen"},{"key":"bibr26-0278364920937608","first-page":"27","author":"Gauss C","year":"1823","journal-title":"Werke, IV, Koniglichen Gesellschaft der Wissenschaften zu Gottingen"},{"key":"bibr27-0278364920937608","first-page":"431","author":"Ghahramani Z","year":"1999","journal-title":"Advances in Neural Information Processing Systems"},{"key":"bibr28-0278364920937608","volume-title":"Matrix Computations","author":"Golub H","year":"1996"},{"issue":"1","key":"bibr29-0278364920937608","first-page":"1303","volume":"14","author":"Hoffman MD","year":"2013","journal-title":"The Journal of Machine Learning Research"},{"key":"bibr30-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1080\/03610927708827533"},{"key":"bibr31-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/9.855552"},{"key":"bibr32-0278364920937608","volume-title":"Stochastic Processes and Filtering Theory","author":"Jazwinski AH","year":"1970"},{"key":"bibr33-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/BF02418571"},{"key":"bibr34-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1016\/j.automatica.2012.11.014"},{"key":"bibr35-0278364920937608","unstructured":"Julier S, Uhlmann J (1996) A general method for approximating nonlinear transformations of probability distributions. Technical report, Robotics Research Group, University of Oxford."},{"key":"bibr36-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1016\/j.robot.2009.06.008"},{"key":"bibr37-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA.2011.5979641"},{"key":"bibr38-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TRO.2008.2006706"},{"key":"bibr39-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1115\/1.3662552"},{"key":"bibr40-0278364920937608","first-page":"1","volume-title":"17th International Conference on Information Fusion (FUSION)","author":"Kokkala J","year":"2014"},{"issue":"1","key":"bibr41-0278364920937608","first-page":"15","volume":"11","author":"Kokkala J","year":"2016","journal-title":"Journal of Advances in Information Fusion"},{"key":"bibr42-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1214\/aoms\/1177729694"},{"key":"bibr43-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2011.2107522"},{"key":"bibr44-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1023\/A:1008854305733"},{"key":"bibr45-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/CRV.2015.52"},{"key":"bibr46-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1002\/9781119541219"},{"key":"bibr47-0278364920937608","author":"Markoff A","year":"1912","journal-title":"Wahrscheinlichheitsrechnung"},{"key":"bibr48-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1177\/0278364917732639"},{"key":"bibr49-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1137\/0613045"},{"key":"bibr50-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1177\/0278364918790369"},{"key":"bibr51-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/978-94-011-5014-9_12"},{"key":"bibr52-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/978-0-387-40065-5_10"},{"key":"bibr53-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1016\/0378-3758(91)90002-V"},{"key":"bibr54-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1162\/neco.2008.08-07-592"},{"key":"bibr55-0278364920937608","volume-title":"Iterative Solution of Nonlinear Equations in Several Variables","author":"Ortega J","year":"1970"},{"key":"bibr56-0278364920937608","doi-asserted-by":"publisher","DOI":"10.5140\/JASS.2009.26.1.075"},{"key":"bibr57-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-28572-1_15"},{"key":"bibr58-0278364920937608","volume-title":"Numerical Recipes: The Art of Scientific Computing","author":"Press WH","year":"2007"},{"key":"bibr59-0278364920937608","first-page":"2191","volume-title":"Proceedings of International Joint Conference on Artificial Intelligence (IJCAI)","author":"Ranganathan A","year":"2007"},{"key":"bibr60-0278364920937608","volume-title":"Gaussian Processes for Machine Learning","author":"Rasmussen CE","year":"2006"},{"key":"bibr61-0278364920937608","doi-asserted-by":"publisher","DOI":"10.2514\/3.3166"},{"issue":"1","key":"bibr62-0278364920937608","first-page":"19","volume":"127","author":"Roh KM","year":"2007","journal-title":"Advances in the Astronautical Sciences"},{"key":"bibr63-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9781139344203"},{"issue":"1","key":"bibr64-0278364920937608","first-page":"31","volume":"11","author":"S\u00e4rkk\u00e4 S","year":"2016","journal-title":"Journal of Advances in Information Fusion"},{"key":"bibr65-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2011.2174667"},{"key":"bibr66-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1016\/j.automatica.2010.10.013"},{"key":"bibr67-0278364920937608","doi-asserted-by":"publisher","DOI":"10.15607\/RSS.2006.II.034"},{"key":"bibr68-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1214\/aos\/1176345632"},{"key":"bibr69-0278364920937608","volume-title":"Proceedings of the PICA Conference","author":"Takahashi K","year":"1973"},{"key":"bibr70-0278364920937608","volume-title":"Probabilistic Robotics","author":"Thrun S","year":"2006"},{"key":"bibr71-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1177\/0278364904045479"},{"issue":"5","key":"bibr72-0278364920937608","first-page":"403","volume":"25","author":"Thrun S","year":"2005","journal-title":"The International Journal on Robotics Research"},{"key":"bibr73-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1007\/3-540-44480-7_21"},{"key":"bibr74-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1177\/0278364906075026"},{"key":"bibr75-0278364920937608","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2006.875389"}],"container-title":["The International Journal of Robotics Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/0278364920937608","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/0278364920937608","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/0278364920937608","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T10:16:17Z","timestamp":1777457777000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/0278364920937608"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,7,29]]},"references-count":75,"journal-issue":{"issue":"13","published-print":{"date-parts":[[2020,11]]}},"alternative-id":["10.1177\/0278364920937608"],"URL":"https:\/\/doi.org\/10.1177\/0278364920937608","relation":{},"ISSN":["0278-3649","1741-3176"],"issn-type":[{"value":"0278-3649","type":"print"},{"value":"1741-3176","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,7,29]]}}}