Abstract:The availability of digital music data in various modalities provides opportunities both for music enjoyment and music research. Regarding the latter, the computer-assisted analysis of tonal structures is a central topic. For Western classical music, studies typically rely on machine-readable scores, which are tedious to create for large-scale works and comprehensive corpora. As an alternative, music audio recordings, which are readily available, can be analyzed with computational methods. With this paper, we want to bridge the gap between score- and audio-based measurements of tonal structures by leveraging the power of deep neural networks. Such networks are commonly trained in an end-to-end fashion, which introduces biases towards the training repertoire or towards specific annotators. To overcome these problems, we propose a multi-step strategy. First, we compute pitch-class representations of the audio recordings using networks trained on score–audio pairs. Second, we measure the presence of specific tonal structures using a pattern-matching technique that solely relies on music theory knowledge and does not require annotated training data. Third, we highlight these measurements with interactive visualizations, thus leaving the interpretation to the musicological experts. Our experiments on Richard Wagner's large-scale cycle Der Ring des Nibelungen indicate that deep pitch-class representations lead to a high similarity between score- and audio-based measurements of tonal structures, thus demonstrating how to leverage multimodal data for application scenarios in the computational humanities, where an explicit and interpretable methodology is essential.

Towards Musicologist-Driven Mining of Handwritten Scores.

Computational music analysis from first principles

Exploring Music21 and Gensim for Music Data Analysis and Visualization

Automatic staff reconstruction within SIMSSA proect

A Deep Learning-Based Piano Music Notation Recognition Method

CorpusVis: Visual Analysis of Digital Sheet Music Collections

Piece Identification in Classical Piano Music Without Reference Scores

Humanities and engineering perspectives on music transcription

Towards Robust and Truly Large-Scale Audio-Sheet Music Retrieval

Low-Data Classification of Historical Music Manuscripts: A Few-Shot Learning Approach

Text and Data Mining for Pianists? Bringing Digital Humanities to a Graduate Music Research Methods Course Through Topic Modeling

BacHMMachine: An Interpretable and Scalable Model for Algorithmic Harmonization for Four-part Baroque Chorales

The Batik-plays-Mozart Corpus: Linking Performance to Score to Musicological Annotations

From Music Scores to Audio Recordings: Deep Pitch-Class Representations for Measuring Tonal Structures

Research on the Multi-non-artistic presentation Form of Western Music History based on Weighted Function Algorithm

BachProp: Learning to Compose Music in Multiple Styles

MusicScore: A Dataset for Music Score Modeling and Generation

Towards Musically Informed Evaluation of Piano Transcription Models

HMM-based writer identification in music score documents without staff-line removal

A classification approach to interpretation of traditional Chinese musical score

Markov-switching State Space Models for Uncovering Musical Interpretation