Single-Channel Speech Enhancement Based on Multi-Scale Gradients Generative Adversarial Networks.
Xiao-Jun ZhuHe-Ming HuangPublished in: CIS (2021)
Keyphrases
- single channel
- speech enhancement
- multi channel
- multiscale
- wavelet decomposition
- noisy environments
- noise reduction
- independent component analysis
- frequency domain
- signal to noise ratio
- prior information
- wiener filter
- brain activity
- edge detection
- multiresolution
- speech signal
- scale space
- image processing
- linear prediction
- wavelet domain
- subband
- sound source
- wavelet transform