A tutorial on particle filters for on-line nonlinear/non-Gaussian Bayesian tracking
read more
Citations
Planning Algorithms: Introductory Material
Kernel-based object tracking
Earthquake shakes Twitter users: real-time event detection by social sensors
A Tutorial on Particle Filtering and Smoothing: Fifteen years later
Gesture Recognition: A Survey
References
Novel approach to nonlinear/non-Gaussian Bayesian state estimation
The viterbi algorithm
On sequential Monte Carlo sampling methods for Bayesian filtering
The unscented Kalman filter for nonlinear estimation
Filtering via Simulation: Auxiliary Particle Filters
Related Papers (5)
Frequently Asked Questions (14)
Q2. What is the way to estimate the state of the particles?
6Since the particles actually represent paths through the state space, by storing the trajectory taken by each particle, fixed-lag and fixed-point smoothed estimates of the state can be obtained [4].
Q3. What is the simplest way to approximate the posterior density?
If the state space is continuous but can be decomposed into “cells,” : , then a grid-based method can be used to approximate the posterior density.
Q4. What is the common problem with the SIS particle filter?
Assign the particle a weight, , according to (48) END FOR1) Degeneracy Problem: A common problem with the SIS particle filter is the degeneracy phenomenon, where after a few iterations, all but one particle will have negligible weight.
Q5. Why is the problem of calculating smoothed densities of data of interest?
The problem of calculating smoothed densities is of interest because the densities at time are then conditional not only on measurements up to and including time index but also on future measurements.
Q6. What is the common approach to calculating the maximum a posteriori estimate of the path?
In HMM-based tracking, a common approach is to use the Viterbi algorithm [18] to calculate the maximum a posteriori estimate of the path through the trellis, that is, the sequence of discrete states that maximizes the probability of the state sequence given the data.
Q7. What is the second method of resampling?
The second method by which the effects of degeneracy can be reduced is to use resampling whenever a significant degeneracy is observed (i.e., when falls below some threshold ).
Q8. How many standard deviations of the mean would be expected?
Were the density to be Gaussian, one would expect the state to be within two standard deviations of the mean approximately 95% of the time.
Q9. What is the importance density used to draw the sample?
The importance density used to draw the sample is defined to satisfy the proportionality(68)where is some characterization of , given .
Q10. What is the basis for particle filters that have been developed so far?
The sequential importance sampling algorithm presented in Section V-A forms the basis for most particle filters that havebeen developed so far.
Q11. What is the ratio of the probability densities of a sample?
Since integrates to unity over while integrates to unity over , the ratio of the probability densities is then proportional to the inverse of the ratio of the lengths, and .
Q12. What is the simplest way to describe the posterior pdf?
In order to develop the details of the algorithm, let denote a random measure that characterizes the posterior pdf , where , is a set of support points with associated weights , and , is the set of all states up to time.
Q13. Why is the ASIR filter used in resampling?
This arises due to the fact that in the resampling stage, samples are drawn from a discrete distribution rather than a continuous one.
Q14. What is the weighted approximation to the density of a sample?
if the samples were drawn from an importance density , then the weights in (40) are defined by (42) to be(43)Returning to the sequential case, at each iteration, one could have samples constituting an approximation to and want to approximate with a new set of samples.