mlsanityData to 7 Oct 2026

Motifs, Phrases, and Beyond: The Modelling of Structure in Symbolic Music Generation

arXiv 2403.07995 ↗First posted 12 Mar 2024cs.SD cs.LG cs.SCcited by 6

Not on the map. Why?

Abstract (arXiv)

Modelling musical structure is vital yet challenging for artificial intelligence systems that generate symbolic music compositions. This literature review dissects the evolution of techniques for incorporating coherent structure, from symbolic approaches to foundational and transformative deep learning methods that harness the power of computation and data across a wide variety of training paradigms. In the later stages, we review an emerging technique which we refer to as "sub-task decomposition" that involves decomposing music generation into separate high-level structural planning and content creation stages. Such systems incorporate some form of musical knowledge or neuro-symbolic methods by extracting melodic skeletons or structural templates to guide the generation. Progress is evident in capturing motifs and repetitions across all three eras reviewed, yet modelling the nuanced development of themes across extended compositions in the style of human composers remains difficult. We outline several key future directions to realize the synergistic benefits of combining approaches from all eras examined.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 34 of its 78 references that are in the map. First posted on arXiv 12 Mar 2024.

Builds on

34 of 78 references.

References · 34

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,530
  3. 2016WaveNet: A Generative Model for Raw Audiocited by 1,710
  4. 2019Generating Long Sequences with Sparse Transformerscited by 1,523
  5. 2015Autoencoding beyond pixels using a learned similarity metriccited by 407
  6. 2018RSA: Byzantine-Robust Stochastic Aggregation Methods for Distributed Learning from Heterogeneous Datasetscited by 148
  7. 2018A Hierarchical Latent Vector Model for Learning Long-Term Structure in Musiccited by 131
  8. 2021Compound Word Transformer: Learning to Compose Full-Song Music over Dynamic Directed Hypergraphscited by 109
  9. 2017MidiNet: A Convolutional Generative Adversarial Network for Symbolic-domain Music Generationcited by 72
  10. 2018This Time with Feeling: Learning Expressive Musical Performancecited by 67
  11. 2019LakhNES: Improving multi-instrumental music generation with cross-domain pre-trainingcited by 54
  12. 2017Deep Learning Techniques for Music Generation -- A Surveycited by 49
22 more
  1. 2022Museformer: Transformer with Fine- and Coarse-Grained Attention for Music Generationcited by 48
  2. 2020SongMASS: Automatic Song Writing with Pre-training and Alignment Constraintcited by 41
  3. 2021Theme Transformer: Symbolic Music Generation with Theme-Conditioned Transformercited by 40
  4. 2022Symphony Generation with Permutation Invariant Language Modelcited by 35
  5. 2021Controllable deep melody generation via hierarchical music structure representationcited by 30
  6. 2018MorpheuS: generating structured music with constrained patterns and tensioncited by 22
  7. 2021AccoMontage: Accompaniment Arrangement via Phrase Selection and Style Transfercited by 22
  8. 2020Automatic Analysis and Influence of Hierarchical Structure on Melody, Rhythm and Harmony in Popular Musiccited by 21
  9. 2021TeleMelody: Lyric-to-Melody Generation with a Template-Based Two-Stage Methodcited by 20
  10. 2021MELONS: generating melody with long-term structure using transformers and structure graphcited by 18
  11. 2022Compose & Embellish: Well-Structured Piano Performance Generation via A Two-Stage Approachcited by 17
  12. 2017A Hierarchical Recurrent Neural Network for Symbolic Melody Generationcited by 14
  13. 2021Hierarchical Recurrent Neural Networks for Conditional Melody Generation with Long-term Structurecited by 12
  14. 2021Structure-Enhanced Pop Music Generation via Harmony-Aware Learningcited by 11
  15. 2023MelodyGLM: Multi-task Pre-training for Symbolic Melody Generationcited by 10
  16. 2022Conditional Drums Generation using Compound Word Representationscited by 7
  17. 2022MeloForm: Generating Melody with Musical Form based on Expert Systems and Neural Networkscited by 5
  18. 2023WuYun: Exploring hierarchical skeleton-guided melody generation using knowledge-enhanced deep learningcited by 5
  19. 2023Pitchclass2vec: Symbolic Music Structure Segmentation with Chord Embeddingscited by 4
  20. 2021Personalized Popular Music Generation Using Imitation and Structurecited by 3
  21. 2022An adaptive music generation architecture for games based on the deep learning Transformer modecited by 3
  22. 2022The Beauty of Repetition in Machine Composition Scenarioscited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 1.

6 papers.

  1. 2025ImprovNet -- Generating Controllable Musical Improvisations with Iterative Corruption Refinementcited by 3
  2. 2025F-StrIPE: Fast Structure-Informed Positional Encoding for Symbolic Music Generationcited by 2
  3. 2025Synthesizing Composite Hierarchical Structure from Symbolic Music Corporacited by 0
  4. 2025Of All StrIPEs: Investigating Structure-informed Positional Encoding for Efficient Music Generationcited by 0
  5. 2025The Ghost in the Keys: A Disklavier Demo for Human-AI Musical Co-Creativitycited by 0
  6. 2026Efficient Long-Sequence Diffusion Modeling for Symbolic Music Generationcited by 0