The Gaussian Mixture Probability Hypothesis Density Filter

By B. Vo, Wing-Kin Ma. IEEE Transactions on Signal Processing, 2006

2,018 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/ba6b51be3d9eb6b3e09a160e63f9eadd703a6fd2

How researchers see this paper

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

In Artificial Intelligence

In Computer Networks and Communications

Citations in the corpus

Multisensor data fusion: A review of the state-of-the-art (2013, methodology): “As a result two approximation methods, namely, Gaussian Mixture (GM) and Sequential Monte Carlo (SMC), have been used in the literature to further ease the implementation stage for these filters [89,90].”
On performance evaluation of multi-object filters (2008, methodology): “The multi-target filters used to test the new metric are the Gaussian mixture PHD [13] and Gaussian mixture CPHD [14] filters.”
The Cardinality Balanced Multi-Target Multi-Bernoulli Filter and Its Implementations (2009, methodology): “For each of the remaining tracks, to reduce the number of components comprising its spatial density, we eliminate components with weights below a threshold T , and merge components within a distance U of each other, analogous to the Gaussian mixture PHD and CPHD filters [11], [12].”
Cubature Kalman Filtering for Continuous-Discrete Systems: Theory and Simulations (2010, neutral): “To handle such complications in a more complicated multitarget multisensor environment, the joint probabilistic data association (JPDA) filter [7], [12], the multihypothesis tracking (MHT) filter [7], [12] and the probability hypothesis density (PHD) filter [34], [ 42 ], [43] are often used.”
Labeled Random Finite Sets and the Bayes Multi-Target Tracking Filter (2013, methodology): “Assuming that the detection probability and single measurement likelihood do not depend on the labels, the updated PHD is given by (see [8], [50], [51]) where the constituent updated PHD due to hypothesis is given by”
Marginal multi-bernoulli filters: RFS derivation of MHT, JIPDA, and association-based member (2012, methodology): “The theorem confirms what one would intuitively expect: that the prediction of the PPP component (37) follows the standard PHD prediction step of [1] [2], and that the multi- Bernoulli tracks are predicted independently in an equivalent manner to the MeMBer [5] [6] (excluding birth of new targets,…”
Labeled Random Finite Sets and Multi-Object Conjugate Priors (2013, background): “Linear Gaussian multi-target model For a linear Gaussian multi-target model [25], which is essentially the assumption of Gaussian mixture single target transition, likelihood and birth intensity, as well as constant survival and detection probabilities, each relevant single target density p k−1(·, `”
A phd Filter for Tracking Multiple Extended Targets Using Random Matrices (2012, methodology): “To keep the number of components at a tractable level, pruning and merging of GIW components is performed similarly to [13].”
Joint Detection and Estimation of Multiple Objects From Image Observations (2010, background): “D. Suter is with the School of Computer Science, University of Adelaide, North Terrace SA 5005, Australia (e-mail: dsuter@cs.adelaide.edu.au).”
Bayesian State-Space Modelling on High-Performance Hardware Using LibBi (2013, background): “biochemistry (Golightly and Wilkinson 2008, 2011) and object tracking (Vo and Ma 2006).”
Tracking of extended objects and group targets using random matrices — a new approach (2008, background): “In particular, the probability hypothesis density (PHD) filter [3]–[5] has been suggested as one possibility to track an unknown number of possibly closely spaced targets while simultaneously solving the association problem in an approximate yet very efficient manner, especially in its Gaussian mixt”
Consensus CPHD Filter for Distributed Multitarget Tracking (2013, methodology): “Merging: Recall [18], [19] that the prediction step (16) and the correction step (17) of the PHD/CPHD filter make the number of Gaussian components increase.”
Adaptive Target Birth Intensity for PHD and CPHD Filters (2012, methodology): “Typically the birth intensity has the majority of its mass distributed over small specific areas of X , which, for example in the air surveillance context, can be interpreted as the regions around airports [10, 13].”
A Consistent Metric for Performance Evaluation of Multi-Object Filters (2008, methodology): “The multi-target filters used to test the new metric are the Gaussian mixture PHD [ 22 ] and Gaussian mixture CPHD [23] filters.”
All Weather Perception: Joint Data Association, Tracking, and Classification for Autonomous Ground Vehicles (2016, methodology): “For the sake of brevity, a full derivation of particle likelihood formulations including birth and death likelihoods and resampling procedures has been omitted; (Vo and Ma, 2006) and (Miller et al., 2011a) both present thorough summaries.”
Deep learning in video multi-object tracking: A survey (2019, methodology): “Afterwards, the matching score was combined with a spatio-temporal relation score, and the final score was used as a likelihood in a Gaussian Mixture Probability Hypothesis Density filter [99].”