Low-Latency Single Channel Speech Enhancement Using U-Net Convolutional Neural Networks.
Ahmet E. BulutKazuhito KoishidaPublished in: ICASSP (2020)
Keyphrases
- low latency
- single channel
- speech enhancement
- convolutional neural networks
- multi channel
- high speed
- high throughput
- highly efficient
- real time
- frequency domain
- virtual machine
- prior information
- wiener filter
- noise reduction
- independent component analysis
- signal to noise ratio
- brain activity
- sound source
- wavelet decomposition
- spatial domain
- speech signal
- noisy environments
- signal processing
- prior knowledge
- feature extraction