A Recurrent Variational Autoencoder for Speech Enhancement.
Simon LeglaiveXavier Alameda-PinedaLaurent GirinRadu HoraudPublished in: ICASSP (2020)
Keyphrases
- speech enhancement
- noise reduction
- noisy environments
- signal to noise ratio
- speech signal
- single channel
- linear prediction
- background noise
- speech recognition
- wiener filter
- smoothing algorithm
- image segmentation
- edge detection
- vocal tract
- optical flow
- multi channel
- additive noise
- random variables
- computer vision
- multiscale