Fully Convolutional Recurrent Networks for Speech Enhancement.
Maximilian StrakeBruno DefraeneKristoff FluytWouter TirryTim FingscheidtPublished in: ICASSP (2020)
Keyphrases
- recurrent networks
- speech enhancement
- recurrent neural networks
- noisy environments
- signal to noise ratio
- feed forward
- noise reduction
- biologically inspired
- single channel
- neural network
- speech signal
- vocal tract
- linear prediction
- smoothing algorithm
- multi channel
- background noise
- artificial neural networks
- additive noise
- speech recognition
- wiener filter
- multiscale
- speech synthesis
- genetic algorithm