Speech Enhancement Based on Reducing the Detail Portion of Speech Spectrograms in Modulation Domain via Discrete Wavelet Transform.
Shih-Kuang LeeSyu-Siang WangYu TsaoJeih-weih HungPublished in: CoRR (2018)
Keyphrases
- discrete wavelet transform
- speech enhancement
- speech signal
- single channel
- noisy environments
- noise reduction
- subband
- wavelet domain
- signal to noise ratio
- wavelet decomposition
- multiresolution
- linear prediction
- wavelet coefficients
- speech recognition
- vocal tract
- high frequency
- automatic speech recognition
- multi channel
- additive noise
- image fusion
- spatial domain
- low frequency
- feature vectors
- wavelet transform
- hidden markov models
- denoising
- speaker identification
- non stationary
- image data
- pattern recognition
- background noise
- wiener filter
- image coding
- bit rate