Multichannel Speech Enhancement Based on Neural Beamforming and a Context-Focused Post-Filtering Network.
Cong PangJingjie FanQifan ShenYue XieChengwei HuangBjörn W. SchullerPublished in: IEEE Trans. Cogn. Dev. Syst. (2024)
Keyphrases
- speech enhancement
- single channel
- linear prediction
- network architecture
- multi channel
- noisy environments
- noise reduction
- signal to noise ratio
- speech signal
- image coding
- independent component analysis
- neural network
- adaptive filtering
- background noise
- lossless compression
- prior information
- prediction error
- spectral analysis
- wiener filter
- smoothing algorithm
- frequency domain