An overview of particle methods for random finite set models

Abstract

This overview paper describes the particle methods developed for the implementation of the a class of Bayes filters formulated using the random finite set formalism. It is primarily intended for the readership already familiar with the particle methods in the context of the standard Bayes filter. The focus in on the Bernoulli particle filter, the probability hypothesis density (PHD) particle filter and the generalised labelled multi-Bernoulli (GLMB) particle filter. The performance of the described filters is demonstrated in the context of bearings-only target tracking application.

Visual end-effector tracking using a 3D model-aided particle filter for humanoid robot platforms

“…The proposed approach paves the way to a number of possible future work comprising: I ) joint estimation of the pose of the end-effector and of the object to manipulate using, e.g., the random Random Finite Set (RFS) filtering approach [49]- [54]; II ) use data fusion techniques to exploit multiple features to improve both performance and robustness of the recursive Bayesian filtering; III ) conduct tests on different robot platform like, e.g., WALK-MAN [2]; IV ) distribute a Free and Open Source Software (FOSS) implementation of the algorithms to the community. …”

A Survey of Recent Advances in Particle Filters and Remaining Challenges for Multitarget Tracking

“…Resampling facilitates online adjusting the number of particles, which is particularly important in the random set PF [ 3 ] where the number of targets is time-varying and in the multi-model PF where the fitness of different models are time-varying [ 39 , 40 ].…”

“…For the general MTT scenes, the FISST developed by Mahler [ 2 ] appears as a powerful tool to develop different algorithms. Correspondingly, there is a large body of extensions of the PF based on FISST, for which the reader is kindly referred to [ 3 ]. In particular, the probability hypothesis density (PHD), which is the intensity associated with the first order moment of the multitarget random finite set (RFS), has been developed as a powerful alternative to the full multitarget posterior for time series recursion [ 137 , 138 ].…”

“…The new theories, algorithms and technologies developed for MTT have further promoted the development of advanced filtering and estimation approaches. For example, with the vitalization of the finite set statistics (FISST) [ 2 ] in the context of MTT, the random set based PF has attracted considerable attention [ 3 ].…”

Distributed multi-sensor multi-view fusion based on generalized covariance intersection

“…The PHD filter can avoid explicit data association and can deal with an unknown and time-varying number of targets. The PHD filter is usually implemented by resorting to either Gaussian Mixtures (GMs) [27] or Sequential Monte Carlo (SMC) approximations [28,29].…”