Toward Interpretable Music Tagging with Self-Attention

TL;DR

This paper introduces a self-attention based deep learning model for music tagging that improves interpretability while maintaining competitive performance, validated on standard datasets and visualized through heat maps.

Contribution

The paper presents a novel self-attention based architecture combining convolutional layers and Transformer encoders for music tagging, enhancing interpretability over traditional models.

Findings

Competitive results on MagnaTagATune and Million Song Dataset

Model's interpretability demonstrated via heat map visualizations

Outperforms fully convolutional and recurrent neural network approaches

Abstract

Self-attention is an attention mechanism that learns a representation by relating different positions in the sequence. The transformer, which is a sequence model solely based on self-attention, and its variants achieved state-of-the-art results in many natural language processing tasks. Since music composes its semantics based on the relations between components in sparse positions, adopting the self-attention mechanism to solve music information retrieval (MIR) problems can be beneficial. Hence, we propose a self-attention based deep sequence model for music tagging. The proposed architecture consists of shallow convolutional layers followed by stacked Transformer encoders. Compared to conventional approaches using fully convolutional or recurrent neural networks, our model is more interpretable while reporting competitive results. We validate the performance of our model with the MagnaTagATune and the Million Song Dataset. In addition, we demonstrate the interpretability of the proposed architecture with a heat map visualization.