A New Approach to Hybrid HMM/ANN Speech Recognition using Mutual Information Neural Networks.
Gerhard RigollChristoph NeukirchenPublished in: NIPS (1996)
Keyphrases
- mutual information
- neural network
- artificial neural networks
- speech recognition
- hidden markov models
- speech signal
- back propagation
- multi stream
- information theoretic
- image registration
- neural nets
- automatic speech recognition systems
- feed forward
- multilayer perceptron
- pattern recognition
- automatic speech recognition
- neural network model
- multi layer perceptron
- speech processing
- feature selection
- recognition engine
- feed forward neural networks
- similarity measure
- recurrent neural networks
- genetic algorithm
- speech synthesis
- hybrid intelligent
- finite state transducers
- information gain
- speaker identification
- medical image registration
- using artificial neural networks
- hidden layer
- nonlinear dynamic systems
- radial basis function
- phoneme recognition
- keyword spotting
- speaker independent
- multi layer
- fuzzy logic
- hybrid models
- conditional mutual information
- mutual information maximization
- speech recognizer
- competitive learning
- audio visual
- fault diagnosis
- text to speech
- ann models
- handwritten word recognition
- activation function
- gesture recognition
- speaker adaptation
- conditional random fields
- genetic algorithm ga
- hierarchical hidden markov model
- hybrid intelligent systems
- noisy environments
- speech enhancement
- fuzzy systems