Compensating Acoustic Mismatch Using Class-Based Histogram Equalization for Robust Speech Recognition.
Youngjoo SuhSungtak KimHoirin KimPublished in: EURASIP J. Adv. Signal Process. (2007)
Keyphrases
- speech recognition
- speech recognition systems
- histogram equalization
- noisy environments
- speech recognizers
- speaker independent
- hidden markov models
- language model
- speech synthesis
- image enhancement
- acoustic models
- speech signal
- speech recognizer
- automatic speech recognition
- contrast enhancement
- pattern recognition
- speaker identification
- gray scale
- gray level
- speech recognition technology
- speech retrieval
- speaker dependent
- noisy speech
- image processing
- transfer function
- image classification
- multiresolution
- bayesian networks