A Recurrent Variational Autoencoder for Speech Enhancement.
Simon LeglaiveXavier Alameda-PinedaLaurent GirinRadu HoraudPublished in: CoRR (2019)
Keyphrases
- speech enhancement
- noisy environments
- noise reduction
- signal to noise ratio
- single channel
- speech signal
- linear prediction
- vocal tract
- image segmentation
- multi channel
- smoothing algorithm
- optical flow
- background noise
- speech recognition
- pattern recognition
- neural network
- edge preserving
- wiener filter
- bayesian networks
- feature extraction