Music Mood Detection Based On Audio And Lyrics With Deep Neural Net

TL;DR

This paper presents a deep learning approach for multimodal music mood prediction using audio and lyrics, outperforming traditional methods on arousal detection and improving valence prediction through optimized modality fusion.

Contribution

It introduces a novel deep learning model for music mood prediction and demonstrates its superiority over classical feature engineering approaches.

Findings

Deep learning outperforms traditional models on arousal detection.

Both approaches perform equally on valence prediction.

Optimized multimodal fusion significantly improves valence prediction.

Abstract

We consider the task of multimodal music mood prediction based on the audio signal and the lyrics of a track. We reproduce the implementation of traditional feature engineering based approaches and propose a new model based on deep learning. We compare the performance of both approaches on a database containing 18,000 tracks with associated valence and arousal values and show that our approach outperforms classical models on the arousal detection task, and that both approaches perform equally on the valence prediction task. We also compare the a posteriori fusion with fusion of modalities optimized simultaneously with each unimodal model, and observe a significant improvement of valence prediction. We release part of our database for comparison purposes.