Weighted variance variational autoencoder for speech enhancement.
Ali GolmakaniMostafa SadeghiXavier Alameda-PinedaRomain SerizelPublished in: CoRR (2022)
Keyphrases
- speech enhancement
- noisy environments
- noise reduction
- signal to noise ratio
- single channel
- linear prediction
- image segmentation
- speech signal
- prediction error
- smoothing algorithm
- multi channel
- wiener filter
- optical flow
- speech recognition
- additive noise
- motion estimation
- background noise
- pattern recognition
- computer vision