# Motif-Centric Representation Learning for Symbolic Music

arXiv: https://arxiv.org/abs/2309.10597 · HTML: https://mlsanity.com/paper/2309.10597/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 37 references that are in the map.
- First posted on arXiv 19 Sep 2023.

## Builds on (most cited first, 11 of 11)

- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning](https://mlsanity.com/paper/2105.04906/) (2021)
- [POP909: A Pop-song Dataset for Music Arrangement Generation](https://mlsanity.com/paper/2008.07142/) (2020)
- [Learning Interpretable Representation for Controllable Polyphonic Music Generation](https://mlsanity.com/paper/2008.07122/) (2020)
- [Theme Transformer: Symbolic Music Generation with Theme-Conditioned Transformer](https://mlsanity.com/paper/2111.04093/) (2021)
- [Deep Music Analogy Via Latent Representation Disentanglement](https://mlsanity.com/paper/1906.03626/) (2019)
- [Controllable deep melody generation via hierarchical music structure representation](https://mlsanity.com/paper/2109.00663/) (2021)
- [PIANOTREE VAE: Structured Representation Learning for Polyphonic Music](https://mlsanity.com/paper/2008.07118/) (2020)
- [MELONS: generating melody with long-term structure using transformers and structure graph](https://mlsanity.com/paper/2110.05020/) (2021)
- [High-Level Control of Drum Track Generation Using Learned Patterns of Rhythmic Interaction](https://mlsanity.com/paper/1908.00948/) (2019)
- [What is missing in deep music generation? A study of repetition and structure in popular music](https://mlsanity.com/paper/2209.00182/) (2022)

## Built on by (most cited first, 1 of 1)

- [Probabilistic Multilabel Graphical Modelling of Motif Transformations in Symbolic Music](https://mlsanity.com/paper/2603.26478/) (2026)

## Abstract (arXiv)

Music motif, as a conceptual building block of composition, is crucial for music structure analysis and automatic composition. While human listeners can identify motifs easily, existing computational models fall short in representing motifs and their developments. The reason is that the nature of motifs is implicit, and the diversity of motif variations extends beyond simple repetitions and modulations. In this study, we aim to learn the implicit relationship between motifs and their variations via representation learning, using the Siamese network architecture and a pretraining and fine-tuning pipeline. A regularization-based method, VICReg, is adopted for pretraining, while contrastive learning is used for fine-tuning. Experimental results on a retrieval-based task show that these two methods complement each other, yielding an improvement of 12.6% in the area under the precision-recall curve. Lastly, we visualize the acquired motif representations, offering an intuitive comprehension of the overall structure of a music piece. As far as we know, this work marks a noteworthy step forward in computational modeling of music motifs. We believe that this work lays the foundations for future applications of motifs in automatic music composition and music information retrieval.