Stochastic Filtering for Motion Trajectory in Image Sequences Using a Monte Carlo Filter with Estimation of Hyper-Parameters.
Naoyuki IchimuraPublished in: ICPR (4) (2002)
Keyphrases
- monte carlo
- motion trajectories
- image sequences
- hyperparameters
- model selection
- cross validation
- video sequences
- bayesian framework
- motion patterns
- bayesian inference
- support vector
- spatio temporal
- random sampling
- prior information
- markov chain
- closed form
- em algorithm
- human motion
- video data
- particle filter
- maximum likelihood
- noise level
- motion analysis
- parameter estimation
- incremental learning
- maximum a posteriori
- posterior distribution
- object tracking
- three dimensional
- optical flow
- missing values
- natural images
- incomplete data
- motion estimation
- machine learning
- motion model
- noisy images
- computer vision
- least squares
- probabilistic model
- optimal solution
- multi frame