Noisy speech recognition using singular value decomposition and two-sided linear prediction.
K. F. WongS. H. LeungH. C. NgPublished in: EUROSPEECH (1993)
Keyphrases
- linear prediction
- singular value decomposition
- speech enhancement
- speech signal
- image coding
- lossless compression
- prediction error
- least squares
- autoregressive
- dimensionality reduction
- vocal tract
- subband
- principal component analysis
- speech recognition
- noisy environments
- spectral analysis
- neural network
- multiscale
- computational complexity
- bayesian networks