Maximum accept and reject (MARS) training of HMM-GMM speech recognition systems.
Vivek TyagiPublished in: INTERSPEECH (2008)
Keyphrases
- speech recognition
- speech recognition systems
- hidden markov models
- speaker identification
- gaussian mixture model
- speaker recognition
- speaker independent
- speech recognizer
- information retrieval
- speech signal
- training set
- pattern recognition
- mixture model
- machine learning
- conditional random fields
- multi modal
- noisy environments
- language model
- mel frequency cepstral coefficients
- face recognition