文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Paper number 648 Bayesian structure from motion
侵权投诉

Paper number 648 Bayesian structure from motion

Abstract:We formulate structure from motion as a Bayesian inference problem, and use a Markov chain Monte Carlo sampler to sample the posterior on this problem. This results in a method that can identify both small and large tracker errors, and yields reco

Wecannowwriteaposteriormodel.Weintroduceasetofdiscretemaskbits,onepermeasurement,inama-trix

;thesemaskbitsdeterminebywhichnoisemodelameasurementisaffected.Amaskbitwillbefora“good”measurement(i.e.oneaffectedbyisotropicGaus-siannoise),andfora“bad”measurement(i.e.onewhichcontainsnoinformationaboutthemodel).Weintroduceaprioron,,whichiszeroformatricesthathavefewerthannon-zeroelementsinsomeroworcolumn,anduniformotherwise;thispriorensuresthatwedonotattemptinferenceforsituationswherewehaveinsuf cientmeasurements.Thelikelihoodisthenproportionalto

constraint

Noticethatthemaximumoftheposteriorcouldwellnotoccuratthemaximumofthelikelihood,becausealthoughthefactorisationmight tthedatawell,thefactormaysatisfythecameraconstraintspoorly.

3.1Samplingtheposterior

Thisformulationcontainsbothadiscreteandacontinu-ouscomponent.Tosamplecontinuousvariables,weuseamethoddueto[6](anddescribedindetailin[19]),whichappendsasetofindependentGaussianrandomvariablesofnoexternalsigni cancetothestatevariablestoobtainaposterioroftheformativelogofthisposteriorhastheformoftheHamiltonian.Theneg-foraparticleinanenergy eld.Wenowusetwotypesofproposalmove:advancetimeforthisparticle;andchoosenewmomenta(whichcanbedonebyGibbssampling,be-causeeachisindependentofeveryother,andofthestatevariables).Thismethodmovestomaximaoftheposterioraboutasfastasgradientdescent,andthensamplesaroundthemaxima.Ifthestateisfarfromamaximum,thenthestatemovesdowntheenergy eld,gatheringmomentum,whichisthenthrownawaybythesecondtypeofmove,sotheparticlewilltendtogettrappedinmaximaandexplorethem.

Fordiscretevariables,wedrawasamplefromthefullconditionalforthatdiscretevariable.Foraposterior

andawithsetofadiscretevariablesvariable(bothdiscretewhichandtakescontinuous)valuesandwhichis

,thefullconditionalonis,givenby:

where

.Sincethisis

aproblemwherethequantityofdataswampsthenum-berofparametersinthemodel,thechoiceof

isfairlyunimportant;themainissueistochoosethevaluetobesmallenoughthatlargetrackererrorsaremaskedalmostcertainly.

The

andthatmaximisearethenobtainedbyasweepalgorithmwhich xes

(resp.)andsolvesthelinearsystemfor

上一页第5页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email