We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: PIANOTREE VAE: Structured Representation Learning for Polyphonic Music

Abstract: The dominant approach for music representation learning involves the deep unsupervised model family variational autoencoder (VAE). However, most, if not all, viable attempts on this problem have largely been limited to monophonic music. Normally composed of richer modality and more complex musical structures, the polyphonic counterpart has yet to be addressed in the context of music representation learning. In this work, we propose the PianoTree VAE, a novel tree-structure extension upon VAE aiming to fit the polyphonic music learning. The experiments prove the validity of the PianoTree VAE via (i)-semantically meaningful latent code for polyphonic segments; (ii)-more satisfiable reconstruction aside of decent geometry learned in the latent space; (iii)-this model's benefits to the variety of the downstream music generation.
Subjects: Audio and Speech Processing (eess.AS); Computation and Language (cs.CL); Machine Learning (cs.LG); Sound (cs.SD)
Journal reference: In Proceedings of 21st International Conference on Music Information Retrieval (ISMIR), Montreal, Canada (virtual conference), 2020
Cite as: arXiv:2008.07118 [eess.AS]
  (or arXiv:2008.07118v1 [eess.AS] for this version)

Submission history

From: Ziyu Wang [view email]
[v1] Mon, 17 Aug 2020 06:48:59 GMT (1986kb,D)

Link back to: arXiv, form interface, contact.