Spectrogram Feature Losses for Music Source Separation
This is an incremental improvement for music source separation systems, potentially benefiting audio processing applications.
The paper tackled music source separation by adding a high-level spectrogram feature loss from a VGG net to the standard pixel-level L2 loss, resulting in improved separation quality for drums and vocals on the musdb18 database.
In this paper we study deep learning-based music source separation, and explore using an alternative loss to the standard spectrogram pixel-level L2 loss for model training. Our main contribution is in demonstrating that adding a high-level feature loss term, extracted from the spectrograms using a VGG net, can improve separation quality vis-a-vis a pure pixel-level loss. We show this improvement in the context of the MMDenseNet, a State-of-the-Art deep learning model for this task, for the extraction of drums and vocal sounds from songs in the musdb18 database, covering a broad range of western music genres. We believe that this finding can be generalized and applied to broader machine learning-based systems in the audio domain.