Non-stationary Environment Compensation Using Sequential EM Algorithm for Robust Speech Recognition.
Haifeng ShenJun GuoGang LiuQunxia LiPublished in: PKDD (2005)
Keyphrases
- non stationary
- speech recognition
- em algorithm
- noisy environments
- expectation maximization
- speech signal
- mixture model
- language model
- maximum likelihood
- automatic speech recognition
- gaussian mixture model
- parameter estimation
- generative model
- speech synthesis
- hidden markov models
- expectation maximisation
- pattern recognition
- speech recognition systems
- probability density function
- speaker identification
- speech recognizer
- probabilistic model
- probabilistic principal component analysis
- noise reduction
- model selection
- multi modal
- feature space
- image sequences
- information retrieval