An application of sequential Monte Carlo samplers: an alternative to particle filters for non-linear non-Gaussian sequential inference with zero process noise

The world is dynamic. There are a vast range of applications which involve real-time processing of data, often derived from multiple sensors, related to uncertain quantities of interest. These Data Fusion and Target Tracking applications are those of interest at this event.
The DFandTT 2012 conference represents a great opportunity for industrialists to hear from, and speak directly with a broad spectrum of their end-users, and give academics working in this field, a platform from which to present their short to near-term future technology developments across video data capture, storage and visualisation.
The event builds on the success of previous events in this series (the most recent of which was in 2008) and Fusion 2010. It aims to complement other conferences with a similar technical focus taking place in 2012 (notably SPIE in USA and Fusion 2012 in Singapore) both geographically and in terms of having a high impact single track event with excellent papers.
Applications of target tracking and/or data fusion are found in diverse civilian and military fields. Civilian applications include air traffic control, navigation, fault tolerant systems and decision problems. In the military domain, applications include surveillance, target identification, command and control, sensor management and weapon guidance.

Simon is the chair for this IET conference and is also technical manager for the 90-strong Intelligence, Surveillance and Reconnaisance team at QinetiQ.

Simon is also one of ten senior QinetiQ fellows, a Fellow of the IET and an honorary research fellow at Imperial. Simon has now been at QinetiQ for about 13 years. During this time, Simon has been very interested in the development and application of next-generation tracking techniques and specifically those that tackle problems often regarded as difficult.

This paper is an example of one such problem, sequential parameter estimation.