Source counting in speech mixtures using a variational EM approach for complex WATSON mixture models.
Lukas DrudeAleksej ChinaevDang Hai Tran VuReinhold Haeb-UmbachPublished in: ICASSP (2014)
Keyphrases
- variational methods
- mixture model
- expectation maximization
- em algorithm
- gaussian mixture model
- unsupervised learning
- density estimation
- generative model
- probabilistic model
- exponential family
- model selection
- gaussian mixture
- language model
- mixture modeling
- finite mixtures
- maximum likelihood
- probability density function
- parameter estimation
- mixture components
- finite mixture models
- automatic model selection
- finite mixture model
- speech signal
- image processing
- computer vision
- model based clustering
- machine learning