Simultaneous recognition of concurrent speech signals using hidden Markov model decomposition.
Andrew VargaRoger K. MoorePublished in: EUROSPEECH (1991)
Keyphrases
- denoising
- hidden markov models
- speech signal
- automatic speech recognition systems
- noisy speech
- speech recognition
- automatic speech recognition
- noisy environments
- gesture recognition
- speech sounds
- speech enhancement
- handwritten word recognition
- background noise
- sequential data
- speaker identification
- object recognition
- linear prediction
- vocal tract
- phoneme recognition
- handwriting recognition
- pattern recognition
- mel frequency cepstral coefficients
- hierarchical hidden markov model
- visual speech
- continuous hidden markov models
- handwritten words
- conditional random fields
- action recognition
- hidden states
- acoustic features
- word recognition
- character recognition
- non stationary
- viterbi algorithm
- spectral subtraction
- multi modal
- feature extraction
- neural network