Score-informed Networks for Music Performance Assessment

TL;DR

This paper introduces three neural network models that incorporate musical score information to improve automatic music performance assessment, demonstrating the benefits of score-informed approaches over traditional score-independent methods.

Contribution

It presents three novel deep learning architectures that integrate score data into performance assessment models, a first in this research area.

Findings

Score-informed models outperform score-independent models.

Different architectures have varying suitability depending on input representation.

Joint embedding and distance matrix models show promising results.

Abstract

The assessment of music performances in most cases takes into account the underlying musical score being performed. While there have been several automatic approaches for objective music performance assessment (MPA) based on extracted features from both the performance audio and the score, deep neural network-based methods incorporating score information into MPA models have not yet been investigated. In this paper, we introduce three different models capable of score-informed performance assessment. These are (i) a convolutional neural network that utilizes a simple time-series input comprising of aligned pitch contours and score, (ii) a joint embedding model which learns a joint latent space for pitch contours and scores, and (iii) a distance matrix-based convolutional neural network which utilizes patterns in the distance matrix between pitch contours and musical score to predict assessment ratings. Our results provide insights into the suitability of different architectures and input representations and demonstrate the benefits of score-informed models as compared to score-independent models.