Time-Domain Audio Source Separation Based on Wave-U-Net Combined with Discrete Wavelet Transform.
Tomohiko NakamuraHiroshi SaruwatariPublished in: ICASSP (2020)
Keyphrases
- discrete wavelet transform
- source separation
- audio signal
- audio features
- wavelet domain
- subband
- multiresolution
- frequency domain
- wavelet coefficients
- denoising
- blind source separation
- image fusion
- low frequency
- high frequency
- independent component analysis
- multimedia
- wavelet transform
- audio visual
- feature vectors
- multiscale
- visual information
- visual features
- signal processing
- feature set
- low level
- simple linear
- music information retrieval
- temporal structure
- wavelet decomposition
- speaker identification
- visual data
- feature selection