Regression-based speech enhancement by convolutional neural network.
Mustafa ErsevenBülent BolatPublished in: SIU (2018)
Keyphrases
- convolutional neural network
- speech enhancement
- noisy environments
- noise reduction
- signal to noise ratio
- face detection
- single channel
- speech signal
- wiener filter
- linear prediction
- multi channel
- background noise
- smoothing algorithm
- vocal tract
- neural network
- edge preserving
- additive noise
- prior information
- object detection
- high resolution
- multiscale