FFC-SE: Fast Fourier Convolution for Speech Enhancement.
Ivan ShchekotovPavel AndreevOleg IvanovAibek AlanovDmitry VetrovPublished in: CoRR (2022)
Keyphrases
- speech enhancement
- fourier domain
- noise reduction
- single channel
- noisy environments
- fast fourier transform
- frequency domain
- signal to noise ratio
- fourier transform
- speech signal
- linear prediction
- spatial domain
- smoothing algorithm
- wiener filter
- vocal tract
- additive noise
- image processing
- wavelet domain
- image reconstruction
- background noise
- information retrieval
- optic flow
- edge detection
- spectral analysis
- point spread function
- feature extraction
- multiscale
- denoising
- blurred images
- edge preserving
- linear combination
- subband
- speech recognition