Shift-Invariant Kernel Additive Modelling for Audio Source Separation

TL;DR

This paper introduces a shift-invariant kernel function for Kernel Additive Modelling in audio source separation, improving robustness to frequency shifts and proposing acceleration techniques to reduce computational costs.

Contribution

It presents a novel shift-invariant kernel for KAM, enhancing separation performance under frequency shifts, along with methods to accelerate the computation.

Findings

Improved separation performance with shift-invariant kernels.

Enhanced robustness to frequency shifts in audio sources.

Reduced computational complexity through acceleration techniques.

Abstract

A major goal in blind source separation to identify and separate sources is to model their inherent characteristics. While most state-of-the-art approaches are supervised methods trained on large datasets, interest in non-data-driven approaches such as Kernel Additive Modelling (KAM) remains high due to their interpretability and adaptability. KAM performs the separation of a given source applying robust statistics on the time-frequency bins selected by a source-specific kernel function, commonly the K-NN function. This choice assumes that the source of interest repeats in both time and frequency. In practice, this assumption does not always hold. Therefore, we introduce a shift-invariant kernel function capable of identifying similar spectral content even under frequency shifts. This way, we can considerably increase the amount of suitable sound material available to the robust statistics. While this leads to an increase in separation performance, a basic formulation, however, is computationally expensive. Therefore, we additionally present acceleration techniques that lower the overall computational complexity.