Phonetic subspace mixture model for speaker diarization.
I-Fan ChenShih-Sian ChengHsin-Min WangPublished in: INTERSPEECH (2010)
Keyphrases
- mixture model
- speaker diarization
- speech recognition
- bayesian information criterion
- language model
- gaussian mixture model
- em algorithm
- generative model
- model selection
- probabilistic model
- expectation maximization
- low dimensional
- maximum likelihood
- speaker identification
- speech signal
- hidden markov models
- high dimensional data
- high dimensional
- unsupervised learning
- dimensionality reduction
- broadcast news
- information retrieval
- noisy environments
- automatic speech recognition
- gaussian mixture
- reinforcement learning
- pattern recognition
- feature space
- machine learning
- feature extraction
- principal component analysis