This talk is concerned with application of Rao-Blackwellized particle filtering to multiple target tracking. In particular, the aim is to discuss formulation of track-based multiple target tracking as Rao-Blackwellized particle filtering in exactly or approximately conditionally Gaussian state-space models. In that formulation, the unknown data-associations and the unknown number of targets act as the non-linear variables which are sampled with sequential Monte Carlo. The aim is also to discuss the corresponding smoothing problem as well as sampling of parameters with particle MCMC methods.

This presentation is part of Minisymposium “MS74 - Sequential Monte Carlo methods for inverse estimation in imaging science”
organized by: Narayan Puthanmadam Subramaniyam (Aalto University) , Sara Sommariva (Aalto University) .

Authors:

Simo Särkkä (Aalto University, Department of Electrical Engineering and Automation)

Keywords:

bayesian methods, bayesian methods, computer vision, multiple target tracking, particle filtering, sequential monte carlo, statistical inverse estimation methods, stochastic processes