Using second-order hidden Markov model to improve speaker identification recognition performance under neutral condition.
Ismail ShahinPublished in: ICECS (2003)
Keyphrases
- hidden markov models
- speaker identification
- speaker independent
- speech recognition
- speech signal
- speaker dependent
- visual speech
- gesture recognition
- handwritten word recognition
- noisy environments
- speaker recognition
- gaussian mixture model
- feature extraction
- speech processing
- speech recognizer
- pattern recognition
- automatic speech recognition
- handwriting recognition
- broadcast news
- unsupervised learning
- mel frequency cepstral coefficients
- computer vision
- maximum likelihood
- handwritten words
- reinforcement learning
- multiscale