English

Tackling Interpretability in Audio Classification Networks with Non-negative Matrix Factorization

Sound 2023-05-15 v1 Machine Learning Audio and Speech Processing

Abstract

This paper tackles two major problem settings for interpretability of audio processing networks, post-hoc and by-design interpretation. For post-hoc interpretation, we aim to interpret decisions of a network in terms of high-level audio objects that are also listenable for the end-user. This is extended to present an inherently interpretable model with high performance. To this end, we propose a novel interpreter design that incorporates non-negative matrix factorization (NMF). In particular, an interpreter is trained to generate a regularized intermediate embedding from hidden layers of a target network, learnt as time-activations of a pre-learnt NMF dictionary. Our methodology allows us to generate intuitive audio-based interpretations that explicitly enhance parts of the input signal most relevant for a network's decision. We demonstrate our method's applicability on a variety of classification tasks, including multi-label data for real-world audio and music.

Keywords

Cite

@article{arxiv.2305.07132,
  title  = {Tackling Interpretability in Audio Classification Networks with Non-negative Matrix Factorization},
  author = {Jayneel Parekh and Sanjeel Parekh and Pavlo Mozharovskyi and Gaël Richard and Florence d'Alché-Buc},
  journal= {arXiv preprint arXiv:2305.07132},
  year   = {2023}
}

Comments

Under submission at IEEE/ACM TASLP. arXiv admin note: text overlap with arXiv:2202.11479

R2 v1 2026-06-28T10:32:29.501Z