audioLIME: Listenable Explanations Using Source Separation
Sound
2020-09-08 v3 Information Retrieval
Machine Learning
Audio and Speech Processing
Abstract
Deep neural networks (DNNs) are successfully applied in a wide variety of music information retrieval (MIR) tasks but their predictions are usually not interpretable. We propose audioLIME, a method based on Local Interpretable Model-agnostic Explanations (LIME) extended by a musical definition of locality. The perturbations used in LIME are created by switching on/off components extracted by source separation which makes our explanations listenable. We validate audioLIME on two different music tagging systems and show that it produces sensible explanations in situations where a competing method cannot.
Cite
@article{arxiv.2008.00582,
title = {audioLIME: Listenable Explanations Using Source Separation},
author = {Verena Haunschmid and Ethan Manilow and Gerhard Widmer},
journal= {arXiv preprint arXiv:2008.00582},
year = {2020}
}
Comments
In The 13th International Workshop on Machine Learning and Music, ECML-PKDD 2020