Time Domain Speech Enhancement With Attentive Multi-scale Approach.
Chen ChenNana HouDuo MaEng Siong ChngPublished in: APSIPA ASC (2021)
Keyphrases
- speech enhancement
- multiscale
- noise reduction
- single channel
- noisy environments
- signal to noise ratio
- frequency domain
- edge detection
- speech signal
- wavelet decomposition
- linear prediction
- vocal tract
- scale space
- wiener filter
- natural images
- denoising
- multi channel
- image representation
- background noise
- image processing
- image segmentation
- visual attention
- edge preserving
- additive noise
- spatial domain
- wavelet transform
- focus of attention
- smoothing algorithm
- multiresolution
- prediction error
- optic flow
- image acquisition
- speech recognition
- hidden markov models