Single channel speech enhancement technique for low SNR quasi-periodic noise based on reduced order linear prediction.
Chandan K. A. ReddyVahid MontazeriYu RaoIssa M. S. PanahiPublished in: GlobalSIP (2015)
Keyphrases
- speech enhancement
- linear prediction
- single channel
- image coding
- speech signal
- prediction error
- lossless compression
- multi channel
- autoregressive
- subband
- noisy environments
- vocal tract
- noise model
- image sequences
- noise reduction
- signal to noise ratio
- speech recognition
- computational complexity
- wavelet domain
- inter frame
- maximum likelihood
- wavelet decomposition
- motion estimation
- multiresolution
- multiscale
- high quality
- face recognition
- feature selection