Convolutional Recurrent MetriCGAN With Spectral Dimension Compression For Full-Band Speech Enhancement.
Zhongshu HouQinwen HuTianchi SunYuxiang HuChangbao ZhuKai ChenPublished in: ICASSP (2023)
Keyphrases
- speech enhancement
- linear prediction
- lossless compression
- noisy environments
- noise reduction
- signal to noise ratio
- compression ratio
- single channel
- image compression
- speech signal
- subband
- image coding
- compression algorithm
- background noise
- smoothing algorithm
- vocal tract
- prediction error
- spectral analysis
- frequency band
- multiresolution
- wiener filter
- hyperspectral
- quadtree
- independent component analysis
- language model