Total least squares based subband modelling for scalable speech representations with damped sinusoids.
Kris HermusWerner VerhelstPatrick WambacqPhilippe LemmerlingPublished in: INTERSPEECH (2000)
Keyphrases
- subband
- total least squares
- filter bank
- wavelet transform
- multiresolution
- image compression
- bit rate
- wavelet coefficients
- feature vectors
- image restoration
- frequency domain
- regularized least squares
- speech signal
- discrete fourier transform
- multiscale
- rate distortion
- interior point methods
- random variables
- signal processing
- image coding
- pairwise