A trust region based optimization for maximum mutual information estimation of HMMS in speech recognition.
Zhi-Jie YanCong LiuYu HuHui JiangPublished in: ICASSP (2009)
Keyphrases
- speech recognition
- mutual information
- hidden markov models
- automatic speech recognition
- speech processing
- image registration
- speech synthesis
- language model
- pattern recognition
- speaker identification
- speech recognizer
- speech recognition systems
- speech signal
- similarity measure
- speech understanding
- noisy environments
- speech recognition technology
- image segmentation
- handwriting recognition
- keyword spotting
- speaker independent
- speaker dependent
- speech retrieval
- non stationary
- multi modal
- feature vectors
- bayesian networks
- computer vision
- acoustic models
- speech recognition errors
- information retrieval