Time and frequency filtering of filter-bank energies for robust HMM speech recognition.
Climent NadeuDusan MachoJavier HernandoPublished in: Speech Commun. (2001)
Keyphrases
- speech recognition
- filter bank
- hidden markov models
- noisy environments
- image coding
- subband
- finite impulse response
- multiresolution
- speech signal
- automatic speech recognition
- computationally efficient
- language model
- frequency domain
- multiscale
- speech recognizer
- pattern recognition
- perfect reconstruction
- speech synthesis
- speech processing
- cepstral coefficients
- speaker identification
- keyword spotting
- low frequency
- signal processing
- speaker independent
- wavelet packet
- speaker adaptation
- speech recognition systems
- low pass
- neural network
- speaker dependent
- wavelet filters
- adaptive filtering
- principal component analysis
- image processing