[DMANET] Postdoctoral position - Bayesian Tracking and Reasoning over Time

A696 - Research Associate - BTaRoT project: Bayesian Tracking and Reasoning over Time

Candidates are invited to apply for a full-time post-doctoral position at Lancaster University, United Kingdom. Lancaster's School of Computing and Communications is a world-class research-led school offering friendly and vibrant environment.

The position is associated with a three-year EPSRC-funded project aimed at the development of sequential Monte Carlo methods for solving problems such as group object tracking and non-parametric inference in sensor networks. The project partners are University of Cambridge and QinetiQ.

Flooded with information, decision making systems have to be able to cope with the deluge of data and hence solve efficiently complex and high dimensional problems. Conventional methods fall short in providing reliable solutions in such cases and a new way of thinking, new methods are needed. This project aims at developing scalable Bayesian approaches able to solve complex and high dimensional problems with multi-sensor data. One such problem is tracking groups and extended objects.

Inference and decision making approaches will be developed based on sequential Monte Carlo methods, parameter learning methods, Markov Chain Monte Carlo (MCMC), combined with compressed sensing for optimal management of the resources of sensor networks. Scenarios with more than hundreds of objects will also be considered. Solutions to these problems could lead to practical implementations and would be directly applicable to a range of multi-sensor multi-target (and group) tracking problems in many sensor applications including transportation systems, radar and wireless networks.

Whilst for a single object tracking there are well established models, modeling the motion of a group of objects is an unresolved question. Between the problems that will be studied are: modeling the interactions within the group components, e.g., within a group of people (such as from video data), of a convoy of vehicles moving in urban environment, and followed by the development of techniques for tracking the motion and the structure of the group based on data coming from a network of sensors (e.g. with radar and LIDAR data). The project will involve also the investigation of extended object tracking, sensor fusion techniques to detect, identify and track formations (collectives) of targets, to provide predictions of future formation behaviour in complex sensor environments and their efficient implementation.

To be considered for this post, you must have a good first degree (minimum class 2:1) and a PhD degree in signal processing, electrical engineering, aerospace engineering, mathematics, statistics, physics, or a related area. The following background will be useful: statistical signal processing, probability theory, Bayesian inference, sequential Monte Carlo methods, optimisation methods or equivalent experience, as well as experience of developing MATLAB software and familiarity with related toolboxes. You will also be experienced in conducting research projects both individually and as part of a team.

Informal enquiries may be directed to Dr Lyudmila Mihaylova, mila.mihaylova@lancaster.ac.uk, tel: +44 (0)1524 510388.

Candidates can apply via the web site http://hr-jobs.lancs.ac.uk/Vacancy.aspx?ref=A696

The closing date is Friday 10 May, 2013 and the interview date Thursday 23 May 2013.
**********************************************************
*
* Contributions to be spread via DMANET are submitted to
*
* DMANET@zpr.uni-koeln.de
*
* Replies to a message carried on DMANET should NOT be
* addressed to DMANET but to the original sender. The
* original sender, however, is invited to prepare an
* update of the replies received and to communicate it
* via DMANET.
*
* DISCRETE MATHEMATICS AND ALGORITHMS NETWORK (DMANET)
* http://www.zaik.uni-koeln.de/AFS/publications/dmanet/
*
**********************************************************