A Deep Neural Network Based Kalman Filter for Time Domain Speech Enhancement.
Hongjiang YuZhiheng OuyangWei-Ping ZhuBenoît ChampagneYunyun JiPublished in: ISCAS (2019)
Keyphrases
- kalman filter
- speech enhancement
- noise reduction
- noisy environments
- single channel
- kalman filtering
- frequency domain
- signal to noise ratio
- speech signal
- neural network
- linear prediction
- particle filter
- vocal tract
- object tracking
- speech recognition
- background noise
- wiener filter
- additive noise
- smoothing algorithm
- mean shift
- automatic speech recognition
- high dimensional
- three dimensional