Efficient Learning of the Parameters of Non-Linear Models Using Differentiable Resampling in Particle Filters.
Conor RosatoLee DevlinVincent BéraudPaul R. HorridgeThomas B. SchönSimon MaskellPublished in: IEEE Trans. Signal Process. (2022)
Keyphrases
- efficient learning
- particle filter
- parameter estimation
- importance sampling
- visual tracking
- dynamical models
- observation model
- likelihood function
- particle filtering
- objective function
- multiple object tracking
- monte carlo
- state space
- kalman filter
- mean shift
- model selection
- state estimation
- learning algorithm
- motion model
- markov chain monte carlo
- expectation maximization
- pattern languages
- markov chain
- markov random field
- search space