Speech enhancement based on undecimated wavelet packet-perceptual filterbanks and MMSE-STSA estimation in various noise environments.
Haci TasmazErgun ErçelebiPublished in: Digit. Signal Process. (2008)
Keyphrases
- wavelet packet
- speech enhancement
- filter bank
- single channel
- wavelet decomposition
- subband
- noisy environments
- linear prediction
- image coding
- noise reduction
- signal to noise ratio
- multiresolution
- wavelet transform
- denoising
- signal processing
- frequency domain
- wiener filter
- discrete wavelet transform
- low frequency
- multiscale
- speech signal
- background noise
- wavelet domain
- high frequency
- wavelet function
- additive noise
- vocal tract
- wavelet coefficients
- speech recognition
- bit rate
- computationally efficient
- image compression
- frequency band
- image denoising
- sar images
- multi channel
- image data
- optical flow
- face recognition
- spatial domain