Expectation Maximization (EM) algorithm-based nonlinear target tracking with adaptive state transition matrix and noise covariance.
Ming LeiChongzhao HanPanzhi LiuPublished in: FUSION (2007)
Keyphrases
- em algorithm
- expectation maximization
- target tracking
- maximum likelihood
- mixture model
- likelihood function
- probabilistic model
- log likelihood function
- gaussian mixture model
- transition matrix
- expectation maximisation
- unsupervised learning
- gaussian mixture
- generative model
- maximum likelihood estimation
- maximum a posteriori
- multi sensor
- image segmentation
- finite mixture model
- data fusion
- hidden variables
- likelihood maximization
- kalman filter
- particle filter
- markov chain
- pairwise
- motion estimation
- density estimation
- object tracking
- mean shift
- random walk
- maximum likelihood estimates
- model based clustering
- mixture modeling
- state space