Unsupervised Music Source Separation Using Differentiable Parametric Source Models

Abstract

Supervised deep learning approaches to underdetermined audio source separation achieve state-of-the-art performance but require a dataset of mixtures along with their corresponding isolated source signals. Such datasets can be extremely costly to obtain for musical mixtures. This raises a need for unsupervised methods. We propose a novel unsupervised model-based deep learning approach to musical source separation. Each source is modelled with a differentiable parametric source-filter model. A neural network is trained to reconstruct the observed mixture as a sum of the sources by estimating the source models' parameters given their fundamental frequencies. At test time, soft masks are obtained from the synthesized source signals. The experimental evaluation on a vocal ensemble separation task shows that the proposed method outperforms learning-free methods based on nonnegative matrix factorization and a supervised deep learning baseline. Integrating domain knowledge in the form of sour

Stats

• citations28
• S2 citations—
• github stars0
• HF likes0
• heat score10.97
• arxiv keyschulzeforster2022unsupervised

Related papers

• Interleaved Multitask Learning For Audio Source Separation With Independent Databases (2019)0.00
• Source Separation And Depthwise Separable Convolutions For Computer Audition (2020)0.00
• Deep Bayesian Unsupervised Source Separation Based On A Complex Gaussian Mixture Model (2019)6.34
• End-to-end Non-negative Autoencoders For Sound Source Separation (2019)2.26
• A Comparison And Combination Of Unsupervised Blind Source Separation Techniques (2021)0.00
• A Recurrent Encoder-decoder Approach With Skip-filtering Connections For Monaural Singing Voice Separation (2017)9.41
• A Style Transfer Approach To Source Separation (2019)3.58
• Surrogate Source Model Learning For Determined Source Separation (2020)9.59