# MusicLIME: Explainable Multimodal Music Understanding

Source: [MusicLIME: Explainable Multimodal Music Understanding](<https://www.sota2.com/research/paper/68fa77771e209131cea734ef>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-27T06:49:55.154Z

Authors: Theodoros Sotirou, Vassilis Lyberatos, Orfeas Menis Mastromichalakis, Giorgos Stamou

Published: 2024-09-16T17:28:21.000Z

## Abstract

Multimodal models are critical for music understanding tasks, as they capture the complex interplay between audio and lyrics. However, as these models become more prevalent, the need for explainability grows-understanding how these systems make decisions is vital for ensuring fairness, reducing bias, and fostering trust. In this paper, we introduce MusicLIME, a model-agnostic feature importance explanation method designed for multimodal music models. Unlike traditional unimodal methods, which analyze each modality separately without considering the interaction between them, often leading to incomplete or misleading explanations, MusicLIME reveals how audio and lyrical features interact and contribute to predictions, providing a holistic view of the model's decision-making. Additionally, we enhance local explanations by aggregating them into global explanations, giving users a broader perspective of model behavior. Through this work, we contribute to improving the interpretability of multimodal music models, empowering users to make informed choices, and fostering more equitable, fair, and transparent music understanding systems.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Multi-class Music Classification on Music4All](<https://www.sota2.com/research/sota/multi-class-music-classification-on-music4all.md>): Multi-class Music Classification; Music4All; Accuracy: 57.34.

- [Multi-class Music Classification on AudioSet](<https://www.sota2.com/research/sota/multi-class-music-classification-on-audioset.md>): Multi-class Music Classification; AudioSet; Accuracy: 58.71.

- [Multi-label Music Auto-tagging on MTG-Jamendo](<https://www.sota2.com/research/sota/multi-label-music-auto-tagging-on-mtg-jamendo.md>): Multi-label Music Auto-tagging; MTG-Jamendo; ROC-AUC: 75.9.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2409.10496>)

[PDF](<https://arxiv.org/pdf/2409.10496.pdf>)