A Constraint-Based Evolutionary Learning Approach to the Expectation Maximization for Optimal Estimation of the Hidden Markov Model for Speech Signal Modeling.
Md. Shamsul HudaJohn YearwoodRoberto TogneriPublished in: IEEE Trans. Syst. Man Cybern. Part B (2009)
Keyphrases
- hidden markov models
- speech signal
- evolutionary learning
- speech recognition
- expectation maximization
- automatic speech recognition
- baum welch
- automatic speech recognition systems
- em algorithm
- background noise
- evolutionary computation
- parameter estimation
- conditional random fields
- noisy environments
- speaker identification
- vocal tract
- reinforcement learning
- optimal solution
- speaker recognition
- noisy speech
- mixture model
- linear prediction
- acoustic features
- search space
- generative model
- gaussian mixture model