Heuristic particle filter: applying abstraction techniques to the design of visual tracking algorithms

Pantrigo, Juan José; Sánchez, Ángel Mediavilla; Montemayor, Antonio S.

doi:10.1111/j.1468-0394.2010.00541.x

Cited by 9 publications

(4 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Building on their work in Pantrigo and Sanchez 61 and Pantrigo et al, 62 Cabido et al 63 show an e®ective approach to visual tracking on a graphical processing unit using a hybridized particle¯lter. They propose a hybridization of the particle¯lter with a memetic algorithm to form the memetic algorithm particle¯lter (MAPF) for tracking single and multiple objects.…”

Section: Hybrid and Metaheuristic Particle¯ltersmentioning

confidence: 98%

State Space Modeling & Bayesian Inference with Computational Intelligence

Hanif

Smith

2015

New Math. and Nat. Computation

View full text Add to dashboard Cite

Recursive Bayesian estimation using sequential Monte Carlos methods is a powerful numerical technique to understand latent dynamics of nonlinear non-Gaussian dynamical systems. It enables us to reason under uncertainty and addresses shortcomings underlying deterministic systems and control theories which do not provide su±cient means of performing analysis and design. In addition, parametric techniques such as the Kalman¯lter and its extensions, though they are computationally e±cient, do not reliably compute states and cannot be used to learn stochastic problems. We review recursive Bayesian estimation using sequential Monte Carlo methods highlighting open problems. Primary of these is the weight degeneracy and sample impoverishment problem. We proceed to detail synergistic computational intelligence sequential Monte Carlo methods which address this. We¯nd that imbuing sequential Monte Carlos with computational intelligence has many advantages when applied to many application and problem domains.

show abstract

Section: Hybrid and Metaheuristic Particle¯ltersmentioning

confidence: 98%

State Space Modeling & Bayesian Inference with Computational Intelligence

Hanif

Smith

2015

New Math. and Nat. Computation

View full text Add to dashboard Cite

show abstract

“…Furthermore, including the LS procedure into the PF framework results in a more flexible algorithm. In fact, the statespace defined in the PF stage can be different from the search space used in the LS stage (Pantrigo et al, 2011). This is very useful in visual tracking problems, and we make use of this feature in the present work.…”

Section: Tracking Modulementioning

confidence: 99%

Human activity recognition based on kinematic features

et al. 2013

View full text Add to dashboard Cite

This paper proposes a new approach to recognize human actions in 2D sequences, based on real-time visual tracking and simple feature extraction of human activities in video sequences. The proposed method emphasizes the simplicity of the strategies used, in an attempt to describe human actions as precisely as other more sophisticated (and more computationally demanding) methods in the literature. Specifically, we propose three complementary modules for the following: (a) tracking; (b) feature extraction; and (c) action recognition. The first module is based on the hybridization of a particle filter and a local search procedure and makes use of a reduced integral image to speed up the weight computation. The feature extraction module characterizes the silhouette of the tracked person by dividing it into rectangular boxes. Then, the system computes statistics on the evolution of these rectangular boxes over time. Finally, the action recognition module passes these statistics to a support vector machine to classify the actions. Experimental results show that the proposed method works in real-time, and its performance is competitive against other state-of-the-art methods.

show abstract

“…The similarities between particle filters and evolutionary algorithms have been known and exploited by many researchers [17]. Patrigo et al [17] recently suggested a framework for combining particle filters with population based metaheuristics for visual articulated motion tracking. Later in 2007, Uosaki et al [20] proposed an evolutionary strategies particle filter (ESP) for the fault detection in dynamic system.…”

Section: Introductionmentioning

confidence: 99%

Hybrid metaheuristic particle filters for stochastic volatility estimation

Smith

Hussain

2012

Proceedings of the 14th Annual Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

In this paper we propose hybrid metaheuristic particle filters for the dual estimation of state and parameters in a stochastic volatility estimation problem. We use evolutionary strategies and real coded genetic algorithms as the metaheuristics. The hybrid metaheuristic particle filters provides more accurate results and uses lesser number of particles for this high dimension estimation problem. We compare the performance of our hybrid algorithms with a sequential importance resampling particle filter (SIR) and the parameter learning algorithm (PLA). Our hybrid particle filters out perform both these algorithms for this particular dual estimation problem.

show abstract

Heuristic particle filter: applying abstraction techniques to the design of visual tracking algorithms

Cited by 9 publications

References 43 publications

State Space Modeling & Bayesian Inference with Computational Intelligence

State Space Modeling & Bayesian Inference with Computational Intelligence

Human activity recognition based on kinematic features

Hybrid metaheuristic particle filters for stochastic volatility estimation

Contact Info

Product

Resources

About