Maximum Discrepancy Generative Regularization and Non-Negative Matrix Factorization for Single Channel Source Separation
This work addresses source separation in audio and image processing, offering an incremental improvement for scenarios with weak supervision.
The paper tackled the problem of single-channel source separation by applying adversarial learning to train generative models, specifically Non-negative Matrix Factorization (NMF) bases, resulting in a clear improvement in reconstructed signals, especially with limited supervision data.
The idea of adversarial learning of regularization functionals has recently been introduced in the wider context of inverse problems. The intuition behind this method is the realization that it is not only necessary to learn the basic features that make up a class of signals one wants to represent, but also, or even more so, which features to avoid in the representation. In this paper, we will apply this approach to the training of generative models, leading to what we call Maximum Discrepancy Generative Regularization. In particular, we apply this to problem of source separation by means of Non-negative Matrix Factorization (NMF) and present a new method for the adversarial training of NMF bases. We show in numerical experiments, both for image and audio separation, that this leads to a clear improvement of the reconstructed signals, in particular in the case where little or no strong supervision data is available.