A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking

Arulampalam, M. Sanjeev; Maskell, Simon; Gordon, Neil; Clapp, Tim

doi:10.1109/78.978374

Cited by 9,766 publications

(7,328 citation statements)

References 35 publications

Supporting

Mentioning

7,209

Contrasting

Unclassified

113

Order By: Relevance

“…The weights are then updated as zt) . In addition, to avoid sample impoverishment, a resampling step need to be regularly applied [Arulampalam et al, 2002].…”

Section: Monte-carlo Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Sampling techniques for audio-visual tracking and head pose estimation

Odobez¹,

Lanz²

2012

Multimodal Signal Processing

View full text Add to dashboard Cite

Analyzing people behaviors in smart environment using multimodal sensors requires to answer a set of typical questions: who are the people, where are they, what activities are they doing, when, with whom are they interacting, and how. In this view, locating people or their faces and characterizing them (e.g. extracting their body or head orientation) allows to address the first two questions (who and where), and is usually one of the first steps before applying higher level multimodal scene analysis algorithms that address the other questions. In the last ten years, tracking algorithms have experienced considerable progresses, particularly in indoor environment or for specific applications, where they have reached a maturity allowing their deployment in real systems and applications. Nevertheless, there are still several issues that can make the tracking difficult: background clutter, potentially small object size; complex shape, appearance, and motion, and their changes over time or across camera views; inaccurate/rough scene calibration or inconsistent camera calibration between views for 3D tracking; real-time processing requirements. In what follows, we discuss some important aspects of tracking algorithms, and ultimately introduce the remaining of the chapter content.Scenarios and Set-ups. Scenarios and application needs strongly influence the considered physical environment, and therefore the set-up (where, how many, and what type of sensors are used) and choice of a tracking 84

show abstract

“…The weights are then updated as zt) . In addition, to avoid sample impoverishment, a resampling step need to be regularly applied [Arulampalam et al, 2002].…”

Section: Monte-carlo Methodsmentioning

confidence: 99%

“…Methods to solve the above equations depend on how these terms are modeled. [Arulampalam et al, 2002] provides a good and detailed review of these. Below, we only discuss two of them.…”

Section: State-space Bayesian Trackingmentioning

confidence: 99%

Sampling techniques for audio-visual tracking and head pose estimation

Odobez¹,

Lanz²

2012

Multimodal Signal Processing

View full text Add to dashboard Cite

show abstract

“…This method consists of two primary stages, i.e. prediction and update [6][7][8][9][10][11][12][13][14]. In the prediction stage, numbers of particles are generated to predict target's location in the current frame from video data.…”

Section: Methodsmentioning

confidence: 99%

Particle Filter With Binary Gaussian Weighting and Support Vector Machine for Human Pose Interpretation

Agustien¹,

Widyanto²,

Endah³

et al. 2010

MST

View full text Add to dashboard Cite

This paper proposes human pose interpretation using particle filter (PF) with Binary Gaussian Weighting and support vector machine (SVM). In the proposed system, particle filter is used to track a human object, then this human object is skeletonized using thinning algorithm and classified using SVM. The classification is to identify human pose, whether it is normal or abnormal behavior. Here particle filter is modified through weight calculation using Gaussian distribution to reduce the computational time. The modified particle filter consists of four main phases. First, particles are generated to predict target's location. Second, the weight of certain particles is calculated and these particles are used to build Gaussian distribution. Third, the weight of all particles is calculated based on Gaussian distribution. Fourth, particles are updated based on each weight. The modified particle filter could reduce computational time of object tracking since this method does not have to calculate particle's weight one by one. To calculate weight, the proposed method builds Gaussian distribution and calculates particle's weight using this distribution. Through an experiment using video data taken in front of the cashier of a convenience store, the proposed method reduced computational time in tracking process until 68.34% in average compared to the conventional one, meanwhile the accuracy of tracking with this new method is comparable with particle filter method, i.e. 90.3%. Combining particle filter with binary Gaussian weighting and support vector machine is promising for advanced early crime scene investigation.

show abstract

“…For linear Gaussian state space models, optimal Bayesian solutions can be obtained by means of the Kalman filter [20] and Rauch-Tung-Striebel smoother [21]. In terms of finite state space models, grid-based methods can be utilised to provide an optimal solution [22,23]. However, a variety of approximate filtering and smoothing methods for nonlinear and infinite state space models exist, which encompass inter alia Gaussian filtering and smoothing methods [19], Monte Carlo Sampling approaches [23] and approximate gridbased methods [24].…”

Section: Bayesian Filtering and Smoothing In Parameter Estimationmentioning

confidence: 99%

“…In approximate grid-based filtering [22,23], the continuous state space is decomposed into pre-defined cells and the filtering distributions are approximated by a weighted sum of δ functions. Discretising the state space into a finite number of states {x j : j = 1, .…”

Section: Approximate Grid-based Filteringmentioning

confidence: 99%