A Multi-Scale Feature Recalibration Network for End-to-End Single Channel Speech Enhancement.
Yang XianYang SunWenwu WangSyed Mohsen NaqviPublished in: IEEE J. Sel. Top. Signal Process. (2021)
Keyphrases
- end to end
- single channel
- speech enhancement
- multi channel
- multiscale
- wavelet decomposition
- noisy environments
- noise reduction
- wiener filter
- independent component analysis
- signal to noise ratio
- frequency domain
- speech signal
- prior information
- brain activity
- edge detection
- face recognition
- image segmentation
- denoising
- linear prediction
- background noise
- feature vectors
- image coding
- wavelet transform
- feature extraction