Multi-object particle filter revisited
MetadataShow full item record
Instead of the filtering density, we are interested in the entire posterior density that describes the random set of object trajectories. So far only Markov Chain Monte Carlo (MCMC) technique have been proposed to approximate the posterior distribution of the set of trajectories. Using labeled random finite set we show how the classical multi-object particle filter (a direct generalisation of the standard particle filter to the multi-object case) can be used to recursively compute posterior distribution of the set of trajectories. The result is a generic Bayesian multi-object tracker that does not require re-computing the posterior at every time step nor running a long Markov chain, and is much more efficient than the MCMC approximations.
Showing items related by title, author, creator and subject.
Vo, Ba Tuong; Vo, Ba-Ngu (2018)© 2018 ISIF This paper extends the generalized labeled multi-Bernoulli (GLMB) tracking filter to a batch multi-target tracker. In a labeled random finite set formulation, a multi-target tracking filter propagates the ...
Vo, Ba-Ngu; Vo, Ba Tuong; Pham, N.; Suter, D. (2010)The problem of jointly detecting multiple objects and estimating their states from image observations is formulated in a Bayesian framework by modeling the collection of states as a random finite set. Analytic characterizations ...
Hoang, Hung; Vo, Ba Tuong (2014)In multi-object stochastic systems, the issue of sensor management is a theoretically and computationally challenging problem. In this paper, we present a novel random finite set (RFS) approach to the multi-target sensor ...