Reliable Accent-Specific Unit Generation With Discriminative Dynamic Gaussian Mixture Selection for Multi-Accent Chinese Speech Recognition.
Chao ZhangYi LiuYunqing XiaXuan WangChin-Hui LeePublished in: IEEE Trans. Speech Audio Process. (2013)
Keyphrases
- speech recognition
- gaussian mixture
- speech synthesis
- hidden markov models
- pattern recognition
- automatic speech recognition
- em algorithm
- language model
- speech recognition technology
- speech recognizer
- speech signal
- speaker identification
- speech processing
- speaker independent
- gaussian mixture model
- closed form
- probability density function
- expectation maximization
- noisy environments
- speech recognition systems
- mixture model
- discriminative training
- machine learning
- n gram
- statistical analysis
- unsupervised learning
- semi supervised
- speaker dependent
- image processing