Parameter sharing in subband likelihood-maximizing beamforming for speech recognition using microphone arrays.
Michael L. SeltzerRichard M. SternPublished in: ICASSP (1) (2004)
Keyphrases
- subband
- frequency domain
- automatic speech recognition
- filter bank
- wavelet transform
- linear array
- linear prediction
- sound source
- image compression
- generalized gaussian distribution
- bit rate
- speech signal
- multiresolution
- wavelet coefficients
- high frequency
- wavelet decomposition
- speech recognition
- speaker diarization
- feature vectors
- discrete wavelet transform
- wavelet domain
- wavelet packet
- locally adaptive
- low frequency
- subband coding
- low pass
- maximum likelihood
- audio visual
- image coding
- image coder
- wavelet coding
- subband decomposition
- non subsampled contourlet transform
- reconstruction error
- hidden markov models
- transform coefficients
- frequency band
- wavelet filters
- visual quality
- high quality
- wavelet subbands
- vector quantization
- signal processing
- multiscale