Multichannel Speech Enhancement Based on Speech Spectral Magnitude Estimation Using Generalized Gamma Prior Distribution.
Tran Huy DatKazuya TakedaFumitada ItakuraPublished in: ICASSP (4) (2006)
Keyphrases
- speech enhancement
- linear prediction
- prior distribution
- single channel
- prior information
- speech signal
- multi channel
- image coding
- posterior distribution
- bayesian framework
- signal to noise ratio
- noisy environments
- noise reduction
- prediction error
- probability distribution
- frequency domain
- bayesian inference
- lossless compression
- vocal tract
- probability density function
- prior knowledge
- generative model
- subband
- parameter estimation
- autoregressive
- spectral analysis
- speech recognition
- maximum a posteriori
- independent component analysis
- wiener filter
- pattern recognition
- automatic speech recognition
- higher order
- edge detection
- wavelet transform
- hyperparameters
- computer vision