Frame-based Maximum a Posteriori Estimation of Second-Order Statistics for Multichannel Speech Enhancement in Presence of Noise.
Julitta BartolewskaKonrad KowalczykPublished in: MMSP (2021)
Keyphrases
- speech enhancement
- maximum a posteriori estimation
- single channel
- multi channel
- gaussian distribution
- noise reduction
- linear prediction
- noisy environments
- signal to noise ratio
- noisy speech
- speech recognition
- wiener filter
- independent component analysis
- speech signal
- frequency domain
- background noise
- image coding
- prior information
- maximum likelihood
- additive noise
- prediction error
- feature extraction
- vocal tract
- smoothing algorithm
- pattern recognition
- multiscale