Noise Spectrum Estimation Based on SNR Discrepancy for Speech Enhancement.
Atanu SahaTetsuya ShimamuraPublished in: IEICE Trans. Inf. Syst. (2014)
Keyphrases
- speech enhancement
- signal to noise ratio
- spectral subtraction
- noise reduction
- noisy environments
- single channel
- speech signal
- linear prediction
- noisy speech
- vocal tract
- multi channel
- background noise
- smoothing algorithm
- wiener filter
- additive noise
- edge detection
- speech recognition
- wavelet decomposition
- edge preserving
- frequency domain
- multiresolution
- automatic speech recognition
- noisy images
- point correspondences
- speech synthesis
- image coding
- signal processing