Depthwise Separable Convolutions Versus Recurrent Neural Networks for Monaural Singing Voice Separation.
Pyry PyykkönenStylianos I. MimilakisKonstantinos DrossosTuomas VirtanenPublished in: MMSP (2020)
Keyphrases
- recurrent neural networks
- feed forward
- neural network
- complex valued
- artificial neural networks
- reservoir computing
- echo state networks
- long short term memory
- cascade correlation
- recurrent networks
- neural model
- hebbian learning
- fourier transform
- feedforward neural networks
- chaotic time series
- nonlinear dynamic systems
- discrete fourier transform
- music information retrieval
- genetic algorithm
- real time