Towards online source counting in speech mixtures applying a variational EM for complex Watson mixture models.
Lukas DrudeAleksej ChinaevDang Hai Tran VuReinhold Haeb-UmbachPublished in: IWAENC (2014)
Keyphrases
- mixture model
- em algorithm
- expectation maximization
- generative model
- gaussian mixture model
- density estimation
- probabilistic model
- maximum likelihood
- unsupervised learning
- gaussian mixture
- model selection
- language model
- finite mixtures
- mixture modeling
- probability density function
- mixture components
- image segmentation
- variational inference
- automatic model selection
- parameter estimation
- exponential family
- speech signal
- maximum a posteriori
- prior knowledge
- face recognition