User-Driven Fine-Tuning for Beat Tracking
Autor(a) principal: | |
---|---|
Data de Publicação: | 2021 |
Outros Autores: | , , |
Tipo de documento: | Artigo |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10316/95810 https://doi.org/10.3390/electronics10131518 |
Resumo: | The extraction of the beat from musical audio signals represents a foundational task in the field of music information retrieval. While great advances in performance have been achieved due the use of deep neural networks, significant shortcomings still remain. In particular, performance is generally much lower on musical content that differs from that which is contained in existing annotated datasets used for neural network training, as well as in the presence of challenging musical conditions such as rubato. In this paper, we positioned our approach to beat tracking from a real-world perspective where an end-user targets very high accuracy on specific music pieces and for which the current state of the art is not effective. To this end, we explored the use of targeted fine-tuning of a state-of-the-art deep neural network based on a very limited temporal region of annotated beat locations. We demonstrated the success of our approach via improved performance across existing annotated datasets and a new annotation-correction approach for evaluation. Furthermore, we highlighted the ability of content-specific fine-tuning to learn both what is and what is not the beat in challenging musical conditions. |
id |
RCAP_68723aa53cdf63ecf68412b2ef8811d7 |
---|---|
oai_identifier_str |
oai:estudogeral.uc.pt:10316/95810 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
User-Driven Fine-Tuning for Beat TrackingBeat trackingTransfer learningUser adaptationThe extraction of the beat from musical audio signals represents a foundational task in the field of music information retrieval. While great advances in performance have been achieved due the use of deep neural networks, significant shortcomings still remain. In particular, performance is generally much lower on musical content that differs from that which is contained in existing annotated datasets used for neural network training, as well as in the presence of challenging musical conditions such as rubato. In this paper, we positioned our approach to beat tracking from a real-world perspective where an end-user targets very high accuracy on specific music pieces and for which the current state of the art is not effective. To this end, we explored the use of targeted fine-tuning of a state-of-the-art deep neural network based on a very limited temporal region of annotated beat locations. We demonstrated the success of our approach via improved performance across existing annotated datasets and a new annotation-correction approach for evaluation. Furthermore, we highlighted the ability of content-specific fine-tuning to learn both what is and what is not the beat in challenging musical conditions.MDPI2021info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articlehttp://hdl.handle.net/10316/95810http://hdl.handle.net/10316/95810https://doi.org/10.3390/electronics10131518eng2079-9292Pinto, AntónioBöck, SebastianCardoso, JaimeDavies, Matthewinfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2022-05-25T01:31:57Zoai:estudogeral.uc.pt:10316/95810Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T21:14:13.683094Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
User-Driven Fine-Tuning for Beat Tracking |
title |
User-Driven Fine-Tuning for Beat Tracking |
spellingShingle |
User-Driven Fine-Tuning for Beat Tracking Pinto, António Beat tracking Transfer learning User adaptation |
title_short |
User-Driven Fine-Tuning for Beat Tracking |
title_full |
User-Driven Fine-Tuning for Beat Tracking |
title_fullStr |
User-Driven Fine-Tuning for Beat Tracking |
title_full_unstemmed |
User-Driven Fine-Tuning for Beat Tracking |
title_sort |
User-Driven Fine-Tuning for Beat Tracking |
author |
Pinto, António |
author_facet |
Pinto, António Böck, Sebastian Cardoso, Jaime Davies, Matthew |
author_role |
author |
author2 |
Böck, Sebastian Cardoso, Jaime Davies, Matthew |
author2_role |
author author author |
dc.contributor.author.fl_str_mv |
Pinto, António Böck, Sebastian Cardoso, Jaime Davies, Matthew |
dc.subject.por.fl_str_mv |
Beat tracking Transfer learning User adaptation |
topic |
Beat tracking Transfer learning User adaptation |
description |
The extraction of the beat from musical audio signals represents a foundational task in the field of music information retrieval. While great advances in performance have been achieved due the use of deep neural networks, significant shortcomings still remain. In particular, performance is generally much lower on musical content that differs from that which is contained in existing annotated datasets used for neural network training, as well as in the presence of challenging musical conditions such as rubato. In this paper, we positioned our approach to beat tracking from a real-world perspective where an end-user targets very high accuracy on specific music pieces and for which the current state of the art is not effective. To this end, we explored the use of targeted fine-tuning of a state-of-the-art deep neural network based on a very limited temporal region of annotated beat locations. We demonstrated the success of our approach via improved performance across existing annotated datasets and a new annotation-correction approach for evaluation. Furthermore, we highlighted the ability of content-specific fine-tuning to learn both what is and what is not the beat in challenging musical conditions. |
publishDate |
2021 |
dc.date.none.fl_str_mv |
2021 |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/article |
format |
article |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10316/95810 http://hdl.handle.net/10316/95810 https://doi.org/10.3390/electronics10131518 |
url |
http://hdl.handle.net/10316/95810 https://doi.org/10.3390/electronics10131518 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
2079-9292 |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
MDPI |
publisher.none.fl_str_mv |
MDPI |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799134039546789888 |