Pre-Training Strategies Using Contrastive Learning and Playlist Information for Music Classification and Similarity
Mostra el registre complet Registre parcial de l'ítem
- dc.contributor.author Alonso-Jiménez, Pablo
- dc.contributor.author Favory, Xavier
- dc.contributor.author Foroughmand, Hadrien
- dc.contributor.author Bourdalas, Grigoris
- dc.contributor.author Serra, Xavier
- dc.contributor.author Lidy, Thomas
- dc.contributor.author Bogdanov, Dmitry
- dc.date.accessioned 2023-04-25T13:42:18Z
- dc.date.available 2023-04-25T13:42:18Z
- dc.date.issued 2023-04-25
- dc.description This work has been accepted at the 2023 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2023), at Rhodes, Greece. June 4-10, 2023.
- dc.description.abstract In this work, we investigate an approach that relies on contrastive learning and music metadata as a weak source of supervision to train music representation models. Recent studies show that contrastive learning can be used with editorial metadata (e.g., artist or album name) to learn audio representations that are useful for different classification tasks. In this paper, we extend this idea to using playlist data as a source of music similarity information and investigate three approaches to generate anchor and positive track pairs. We evaluate these approaches by fine-tuning the pre-trained models for music multi-label classification tasks (genre, mood, and instrument tagging) and music similarity. We find that creating anchor and positive track pairs by relying on cooccurrences in playlists provides better music similarity and competitive classification results compared to choosing tracks from the same artist as in previous works. Additionally, our best pre-training approach based on playlists provides superior classification performance for most datasets.ca
- dc.description.sponsorship This research was partially funded by Musical AI - PID2019-111403GBI00/ AEI/10.13039/501100011033 of the Spanish Ministerio de Ciencia, Innovacion y Universidades
- dc.format.mimetype application/pdf*
- dc.identifier.uri http://hdl.handle.net/10230/56564
- dc.language.iso engca
- dc.relation.projectID info:eu-repo/grantAgreement/ES/2PE/PID2019-111403GBI00
- dc.rights.accessRights info:eu-repo/semantics/openAccessca
- dc.subject.keyword music representation learning
- dc.subject.keyword contrastive learning
- dc.subject.keyword music classification
- dc.subject.keyword music similarity
- dc.subject.keyword pre-training neural networks
- dc.title Pre-Training Strategies Using Contrastive Learning and Playlist Information for Music Classification and Similarityca
- dc.type info:eu-repo/semantics/preprintca
- dc.type.version info:eu-repo/semantics/submittedVersionca