Feature extraction for robust speech recognition based on maximizing the sharpness of the power distribution and on power flooring.
Chanwoo KimRichard M. SternPublished in: ICASSP (2010)
Keyphrases
- speech recognition
- power distribution
- feature extraction
- noisy environments
- pattern recognition
- speaker identification
- transmission line
- automatic speech recognition
- cepstral coefficients
- hidden markov models
- speech processing
- speech recognition technology
- speech synthesis
- speech recognizer
- language model
- speech recognition systems
- speaker recognition
- speech signal
- speaker independent
- speaker dependent
- feature selection
- wavelet transform
- image processing
- extracted features
- probabilistic model
- feature vectors
- feature space
- neural network
- speech recognizers