Evaluating Disentangled Representations for Controllable Music Generation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914330951286784 |
|---|---|
| author | Ibáñez-Martínez, Laura Nkama, Chukwuemeka Poltronieri, Andrea Serra, Xavier Rocamora, Martín |
| author_facet | Ibáñez-Martínez, Laura Nkama, Chukwuemeka Poltronieri, Andrea Serra, Xavier Rocamora, Martín |
| contents | Recent approaches in music generation rely on disentangled representations, often labeled as structure and timbre or local and global, to enable controllable synthesis. Yet the underlying properties of these embeddings remain underexplored. In this work, we evaluate such disentangled representations in a set of music audio models for controllable generation using a probing-based framework that goes beyond standard downstream tasks. The selected models reflect diverse unsupervised disentanglement strategies, including inductive biases, data augmentations, adversarial objectives, and staged training procedures. We further isolate specific strategies to analyze their effect. Our analysis spans four key axes: informativeness, equivariance, invariance, and disentanglement, which are assessed across datasets, tasks, and controlled transformations. Our findings reveal inconsistencies between intended and actual semantics of the embeddings, suggesting that current strategies fall short of producing truly disentangled representations, and prompting a re-examination of how controllability is approached in music generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_10058 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Evaluating Disentangled Representations for Controllable Music Generation Ibáñez-Martínez, Laura Nkama, Chukwuemeka Poltronieri, Andrea Serra, Xavier Rocamora, Martín Sound Machine Learning Audio and Speech Processing Recent approaches in music generation rely on disentangled representations, often labeled as structure and timbre or local and global, to enable controllable synthesis. Yet the underlying properties of these embeddings remain underexplored. In this work, we evaluate such disentangled representations in a set of music audio models for controllable generation using a probing-based framework that goes beyond standard downstream tasks. The selected models reflect diverse unsupervised disentanglement strategies, including inductive biases, data augmentations, adversarial objectives, and staged training procedures. We further isolate specific strategies to analyze their effect. Our analysis spans four key axes: informativeness, equivariance, invariance, and disentanglement, which are assessed across datasets, tasks, and controlled transformations. Our findings reveal inconsistencies between intended and actual semantics of the embeddings, suggesting that current strategies fall short of producing truly disentangled representations, and prompting a re-examination of how controllability is approached in music generation. |
| title | Evaluating Disentangled Representations for Controllable Music Generation |
| topic | Sound Machine Learning Audio and Speech Processing |
| url | https://arxiv.org/abs/2602.10058 |