A hybrid segmental neural net/hidden Markov model system for continuous speech recognition.
George ZavaliagkosYing ZhaoRichard M. SchwartzJohn MakhoulPublished in: IEEE Trans. Speech Audio Process. (1994)
Keyphrases
- hidden markov models
- neural nets
- continuous speech recognition
- feed forward
- neural network
- back propagation
- speech recognition
- counter propagation
- artificial neural networks
- sequential data
- markov models
- conditional random fields
- markov model
- continuous hidden markov models
- viterbi algorithm
- state transition
- learning tasks
- hidden state
- hidden states
- baum welch
- dynamic probabilistic networks
- speech signal
- handwritten word recognition
- knn
- learning algorithm
- genetic algorithm