Unsupervised Speech Enhancement Using Dynamical Variational Autoencoders.
Xiaoyu BieSimon LeglaiveXavier Alameda-PinedaLaurent GirinPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2022)
Keyphrases
- speech enhancement
- noisy environments
- noise reduction
- denoising
- signal to noise ratio
- single channel
- wiener filter
- speech signal
- image segmentation
- vocal tract
- linear prediction
- smoothing algorithm
- background noise
- additive noise
- multi channel
- neural network
- image acquisition
- noisy images
- image processing
- edge preserving
- image denoising
- restricted boltzmann machine
- random variables
- optical flow
- multiscale