Parameter sharing and minimum classification error training of mixtures of factor analyzers for speaker identification.
Hiroyoshi YamamotoYoshihiko NankakuChiyomi MiyajimaKeiichi TokudaTadashi KitamuraPublished in: ICASSP (1) (2004)
Keyphrases
- factor analyzers
- minimum classification error
- speaker identification
- gaussian mixture
- gaussian mixture model
- statistical pattern recognition
- discriminative training
- speech recognition
- maximum likelihood estimation
- low dimensional
- hidden markov models
- classification error
- feature extraction
- speech signal
- noisy environments
- expectation maximization
- em algorithm
- broadcast news
- training samples
- data sets
- training set
- decision trees
- pattern recognition
- semi supervised
- natural language processing
- maximum likelihood