A signal subspace approach to spatio-temporal prediction for multichannel speech enhancement.
Adam BorowiczPublished in: EURASIP J. Audio Speech Music. Process. (2015)
Keyphrases
- speech enhancement
- linear prediction
- single channel
- signal subspace
- prediction error
- image coding
- speech signal
- multi channel
- noise reduction
- lossless compression
- sensor array
- correlation matrix
- linear array
- vocal tract
- wavelet decomposition
- subband
- noisy environments
- image sequences
- speech recognition
- autoregressive
- frequency domain
- signal to noise ratio
- noise removal
- wiener filter
- image data
- image segmentation
- synthetic aperture radar
- multiresolution