Fullsubnet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement.
Xiang HaoXiangdong SuRadu HoraudXiaofei LiPublished in: ICASSP (2021)
Keyphrases
- single channel
- speech enhancement
- wavelet decomposition
- subband
- frequency domain
- multi channel
- wavelet transform
- multiresolution
- linear prediction
- wavelet coefficients
- filter bank
- wavelet domain
- high frequency
- feature vectors
- multiscale
- image compression
- bit rate
- noise reduction
- noisy environments
- spatial domain
- independent component analysis
- signal to noise ratio
- feature extraction
- brain activity
- image segmentation
- control system
- prior information
- image coding
- noisy images
- feature selection
- denoising
- signal processing