Adapting the Sample Size in Particle Filters Through KLD-Sampling

By Dieter Fox. Int. J. Robotics Res., 2003

809 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Artificial Intelligence, Ocean Engineering

View Paper: https://www.semanticscholar.org/paper/4bee59c548199aa3931be56550a145028cb950c2

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Artificial Intelligence

In Ocean Engineering

Citations in the corpus

Adaptive methods for sequential importance sampling with application to state space models (2008, methodology): “, by increasing the number of particle s until the total weight mass reaches a positive threshold, see [12], or until the Kullback-Leibler divergence(KLD) between the true and estimated target distributions is below a given threshold, see [7].”
Bayesian Filtering for Location Estimation (2003, background): “MHT represents the belief as mixtures of Gaussians: 4 ( 5 )”
A Survey on Active Simultaneous Localization and Mapping: State of the Art and New Frontiers (2022, background): “with p1(x) and p2(x) the prior and posterior distributions (as in MI) [141], or the estimated and true posteriors assuming the latter can be somehow approximated [42], [142], [143].”
Resampling Methods for Particle Filtering: Classification, implementation, and strategies (2015, methodology): “This is the core rationale of the Kullback–Leibler divergence (KLD)-sampling approach [26], which determines the needed number of particles based on the KLD between the samplebased maximum likelihood estimate and a distribution of interest.”
Pedestrian localisation for indoor environments (2008, methodology): “KLD-sampling assumes that the sample-based representation of the propagated belief can be used as an estimate for the posterior [6], and that the true posterior can be represented by a discrete piecewise constant distribution consisting of a set of multidimensional bins.”
Object disappearance for object discovery (2012, background): “Robot localization is a well-studied problem; in indoor environments with good sensors, it is effectively solved by AMCL [19].”
Distributed Multirobot Exploration and Mapping (2005, background): “Particle filters have been applied with great success to this problem [7], [8], [15], [21].”
Point-Based Value Iteration for Continuous POMDPs (2006, background): “Results are averaged for 10 repetitions and the bars represent the standard deviation. with a few-components Gaussian mixture, when the action model is not linear, or when using an on-line mechanism to dynamically adjust the number of particles (Fox, 2002, 2003).”