Examining the Perceptual Effect of Alternative Objective Functions for Deep Learning Based Music Source Separation
- 1 October 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 679-683
Abstract
In this study, we examine the effect of various objective functions used to optimize the recently proposed deep learning architecture for singing voice separation MaD - Masker and Denoiser. The parameters of the MaD architecture are optimized using an objective function that contains a reconstruction criterion between predicted and true magnitude spectra of the singing voice, and a regularization term. We examine various reconstruction criteria such as the generalized Kullback-Leibler, mean squared error, and noise to mask ratio. We also explore recently proposed, for optimizing MaD, regularization terms such as sparsity and TwinNetwork regularization. Results from both objective assessment and listening tests suggest that the TwinNetwork regularization results in improved singing voice separation quality.Keywords
This publication has 17 references indexed in Scilit:
- Deep Neural Network Based Speech Separation Optimizing an Objective Estimator of Intelligibility for Low Latency Applications2018
- Towards Complex Nonnegative Matrix Factorization with the Beta-Divergence2018
- MaD TwinNet: Masker-Denoiser Architecture with Twin Networks for Monaural Sound Source Separation2018
- The 2018 Signal Separation Evaluation Campaign2018
- Alternative Objective Functions for Deep Clustering2018
- The Dimensions of Perceptual Quality of Sound Source Separation2018
- Evaluation of quality of sound source separation algorithms: Human perception vs quantitative metrics2016
- Noise-to-mask ratio minimization by weighted non-negative matrix factorization2010
- Finding structure in audio for music information retrievalIEEE Signal Processing Magazine, 2006
- RASTA processing of speechIEEE Transactions on Speech and Audio Processing, 1994