mlsanityData to 7 Oct 2026

Frechet Music Distance: A Metric For Generative Symbolic Music Evaluation

arXiv 2412.07948 ↗First posted 10 Dec 2024cs.SD cs.AI cs.MMcited by 12

Not on the map. Why?

Abstract (arXiv)

In this paper we introduce the Frechet Music Distance (FMD), a novel evaluation metric for generative symbolic music models, inspired by the Frechet Inception Distance (FID) in computer vision and Frechet Audio Distance (FAD) in generative audio. FMD calculates the distance between distributions of reference and generated symbolic music embeddings, capturing abstract musical features. We validate FMD across several datasets and models. Results indicate that FMD effectively differentiates model quality, providing a domain-specific metric for evaluating symbolic music generation, and establishing a reproducible standard for future research in symbolic music modeling.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 15 of its 29 references that are in the map. First posted on arXiv 10 Dec 2024.

Builds on

15 of 29 references.

References · 15

  1. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  2. 2018Fr'echet Audio Distance: A Metric for Evaluating Music Enhancement Algorithmscited by 221
  3. 2018Enabling Factorized Piano Music Modeling and Generation with the MAESTRO Datasetcited by 215
  4. 2017MuseGAN: Multi-track Sequential Generative Adversarial Networks for Symbolic Music Generation and Accompanimentcited by 151
  5. 2016C-RNN-GAN: Continuous recurrent neural networks with adversarial trainingcited by 132
  6. 2023Adapting Frechet Audio Distance for Generative Music Evaluationcited by 92
  7. 2020POP909: A Pop-song Dataset for Music Arrangement Generationcited by 83
  8. 2020The Jazz Transformer on the Front Line: Exploring the Shortcomings of AI-composed Music through Quantitative Measurescited by 53
  9. 2021Theme Transformer: Symbolic Music Generation with Theme-Conditioned Transformercited by 40
  10. 2020MusPy: A Toolkit for Symbolic Music Generationcited by 37
  11. 2016Music transcription modelling and composition using deep learningcited by 36
  12. 2024MidiCaps: A large-scale MIDI dataset with text captionscited by 19
3 more
  1. 2024Correlation of Fr'echet Audio Distance With Human Perception of Environmental Audio Is Embedding Dependantcited by 15
  2. 2018A database linking piano and orchestral MIDI scores with application to automatic projective orchestrationcited by 11
  3. 2024Symbotunes: unified hub for symbolic music generative modelscited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 6; 2026: 6.

12 papers.

  1. 2025NotaGen: Advancing Musicality in Symbolic Music Generation with Large Language Model Training Paradigmscited by 29
  2. 2025Aligning Text-to-Music Evaluation with Human Preferencescited by 20
  3. 2025Survey on the Evaluation of Generative Models in Musiccited by 10
  4. 2025Unveiling Audio Deepfake Origins: A Deep Metric learning And Conformer Network Approach With Ensemble Fusioncited by 5
  5. 2025PhraseVAE and PhraseLDM: Latent Diffusion for Full-Song Multitrack Symbolic Music Generationcited by 2
  6. 2026BEAT: Tokenizing and Generating Symbolic Music by Uniform Temporal Stepscited by 2
  7. 2026How Far Can Pretrained LLMs Go in Symbolic Music? Controlled Comparisons of Supervised and Preference-based Adaptationcited by 1
  8. 2026BeatEdit: Symbolic Music Generation as Explicit Editingcited by 1
  9. 2025Conditional Diffusion as Latent Constraints for Controllable Symbolic Music Generationcited by 0
  10. 2026Can LLMs understand LilyPond? A benchmark for symbolic music generation and understandingcited by 0
  11. 2026ChartGenEval: Corruption-Tested Multi-Dimensional Feedback for Rhythm-Game Chart Generationcited by 0
  12. 2026Integrating Contextual Embeddings into Evaluation of Expressive MIDI Piano Performancescited by 0