Combating Reverberation in NTF-Based Speech Separation Using a Sub-Source Weighted Multichannel Wiener Filter and Linear Prediction.
Mieszko FrasMarcin WitkowskiKonrad KowalczykPublished in: Interspeech (2021)
Keyphrases
- linear prediction
- speech enhancement
- wiener filter
- speech signal
- sound source
- additive noise
- image coding
- vocal tract
- prediction error
- lossless compression
- single channel
- speech recognition
- image restoration
- denoising
- noise reduction
- point spread function
- autoregressive
- subband
- signal to noise ratio
- minimum variance
- multiscale
- blurred images
- similarity measure
- non stationary
- hidden markov models