Unsupervised speaker indexing using speaker model selection based on Bayesian information criterion.
Masafumi NishidaTatsuya KawaharaPublished in: ICASSP (1) (2003)
Keyphrases
- bayesian information criterion
- model selection
- speaker diarization
- information criterion
- cross validation
- hyperparameters
- unsupervised learning
- mixture model
- sample size
- machine learning
- regression model
- parameter estimation
- gaussian mixture model
- selection criterion
- variable selection
- gaussian process
- statistical inference
- marginal likelihood
- supervised learning
- speaker verification
- feature selection
- model selection criteria
- speech recognition
- semi supervised
- akaike information criterion
- objective function
- bayesian networks
- automatic model selection