Distributed Navigation with Unknown Initial Poses and Data Association via Expectation Maximization

Indelman, Vadim; Michael, Nathan; Dellaert, Frank

Title:

Distributed Navigation with Unknown Initial Poses and Data Association via Expectation Maximization

dc.contributor.author	Indelman, Vadim
dc.contributor.author	Michael, Nathan
dc.contributor.author	Dellaert, Frank
dc.contributor.corporatename	Georgia Institute of Technology. Institute for Robotics and Intelligent Machines	en_US
dc.contributor.corporatename	Georgia Institute of Technology. College of Computing	en_US
dc.contributor.corporatename	Georgia Institute of Technology. School of Interactive Computing	en_US
dc.contributor.corporatename	Carnegie-Mellon University. Robotics Institute	en_US
dc.contributor.corporatename	Technion - Israel Institute of Technology. Faculty of Aerospace Engineering	en_US
dc.contributor.corporatename	Ṭekhniyon, Makhon ṭekhnologi le-Yiśraʾel. Faculty of Aerospace Engineering	en_US
dc.date.accessioned	2016-05-09T19:49:37Z
dc.date.available	2016-05-09T19:49:37Z
dc.date.issued	2015-02
dc.description.abstract	We present a novel approach for multi-robot distributed and incremental inference over variables of interest, such as robot trajectories, considering the initial relative poses between the robots and multi-robot data association are both unknown. Assuming robots share with each other informative observations, this inference problem is formulated within an Expectation-Maximization (EM) optimization, performed by each robot separately, alternating between inference over variables of interest and multi-robot data association. To facilitate this process, a common reference frame between the robots should first be established. We show the latter is coupled with determining multi-robot data association, and therefore concurrently infer both using a separate EM optimization. This optimization is performed by each robot starting from several promising initial solutions, converging to locally-optimal hypotheses regarding data association and reference frame transformation. Choosing the best hypothesis in an incremental problem setting is in particular challenging due to high sensitivity to measurement aliasing and possibly insufficient amount of data. Selecting an incorrect hypothesis introduces outliers and can lead to catastrophic results. To address these challenges we develop a model-selection based approach to choose the most probable hypothesis, while resorting to Chinese Restaurant Process to represent statistical knowledge regarding hypothesis prior probabilities. We evaluate our approach in real-data experiments.	en_US
dc.embargo.terms	null	en_US
dc.identifier.citation	Indelman, V.; Nelson, E.; Michael N.; & Dellaert, F. (2015). Distributed Navigation with Unknown Initial Poses and Data Association via Expectation Maximization. 55th Israel Annual Conference on Aerospace Sciences, February 2015.	en_US
dc.identifier.uri	http://hdl.handle.net/1853/54788
dc.language.iso	en_US	en_US
dc.publisher	Georgia Institute of Technology	en_US
dc.subject	Fault detection and isolation	en_US
dc.subject	Guidance navigation	en_US
dc.subject	Parameter estimation	en_US
dc.title	Distributed Navigation with Unknown Initial Poses and Data Association via Expectation Maximization	en_US
dc.type	Text
dc.type.genre	Proceedings
dspace.entity.type	Publication
local.contributor.author	Dellaert, Frank
local.contributor.corporatename	Institute for Robotics and Intelligent Machines (IRIM)
local.contributor.corporatename	College of Computing
relation.isAuthorOfPublication	dac80074-d9d8-4358-b6eb-397d95bdc868
relation.isOrgUnitOfPublication	66259949-abfd-45c2-9dcc-5a6f2c013bcf
relation.isOrgUnitOfPublication	c8892b3c-8db6-4b7b-a33a-1b67f7db2021