Using generic summarization to improve music information retrieval tasks
Autor(a) principal: | |
---|---|
Data de Publicação: | 2016 |
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/10071/12269 |
Resumo: | In order to satisfy processing time constraints, many music information retrieval (MIR) tasks process only a segment of the whole music signal. This may lead to decreasing performance, as the most important information for the tasks may not be in the processed segments. We leverage generic summarization algorithms, previously applied to text and speech, to summarize items in music datasets. These algorithms build summaries (both concise and diverse), by selecting appropriate segments from the input signal, also making them good candidates to summarize music. We evaluate the summarization process on binary and multiclass music genre classification tasks, by comparing the accuracy when using summarized datasets against the accuracy when using human-oriented summaries, continuous segments (the traditional method used for addressing the previously mentioned time constraints), and full songs of the original dataset. We show that GRASSHOPPER, LexRank, LSA, MMR, and a Support Sets-based centrality model improve classification performance when compared to selected baselines. We also show that summarized datasets lead to a classification performance whose difference is not statistically significant from using full songs. Furthermore, we make an argument stating the advantages of sharing summarized datasets for future MIR research. |
id |
RCAP_423584ee41244d08a81a8854ac98204e |
---|---|
oai_identifier_str |
oai:repositorio.iscte-iul.pt:10071/12269 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Using generic summarization to improve music information retrieval tasksAutomatic music summarizationGeneric summarization algorithmsMusic classificationIn order to satisfy processing time constraints, many music information retrieval (MIR) tasks process only a segment of the whole music signal. This may lead to decreasing performance, as the most important information for the tasks may not be in the processed segments. We leverage generic summarization algorithms, previously applied to text and speech, to summarize items in music datasets. These algorithms build summaries (both concise and diverse), by selecting appropriate segments from the input signal, also making them good candidates to summarize music. We evaluate the summarization process on binary and multiclass music genre classification tasks, by comparing the accuracy when using summarized datasets against the accuracy when using human-oriented summaries, continuous segments (the traditional method used for addressing the previously mentioned time constraints), and full songs of the original dataset. We show that GRASSHOPPER, LexRank, LSA, MMR, and a Support Sets-based centrality model improve classification performance when compared to selected baselines. We also show that summarized datasets lead to a classification performance whose difference is not statistically significant from using full songs. Furthermore, we make an argument stating the advantages of sharing summarized datasets for future MIR research.IEEE2016-12-15T11:18:16Z2016-01-01T00:00:00Z20162019-04-10T10:07:41Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfhttp://hdl.handle.net/10071/12269eng2329-929010.1109/TASLP.2016.2541299Raposo, F.Ribeiro, R.de Matos, D. M.info: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:RCAAP2023-11-09T17:40:14Zoai:repositorio.iscte-iul.pt:10071/12269Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T22:18:36.490616Repositório 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 |
Using generic summarization to improve music information retrieval tasks |
title |
Using generic summarization to improve music information retrieval tasks |
spellingShingle |
Using generic summarization to improve music information retrieval tasks Raposo, F. Automatic music summarization Generic summarization algorithms Music classification |
title_short |
Using generic summarization to improve music information retrieval tasks |
title_full |
Using generic summarization to improve music information retrieval tasks |
title_fullStr |
Using generic summarization to improve music information retrieval tasks |
title_full_unstemmed |
Using generic summarization to improve music information retrieval tasks |
title_sort |
Using generic summarization to improve music information retrieval tasks |
author |
Raposo, F. |
author_facet |
Raposo, F. Ribeiro, R. de Matos, D. M. |
author_role |
author |
author2 |
Ribeiro, R. de Matos, D. M. |
author2_role |
author author |
dc.contributor.author.fl_str_mv |
Raposo, F. Ribeiro, R. de Matos, D. M. |
dc.subject.por.fl_str_mv |
Automatic music summarization Generic summarization algorithms Music classification |
topic |
Automatic music summarization Generic summarization algorithms Music classification |
description |
In order to satisfy processing time constraints, many music information retrieval (MIR) tasks process only a segment of the whole music signal. This may lead to decreasing performance, as the most important information for the tasks may not be in the processed segments. We leverage generic summarization algorithms, previously applied to text and speech, to summarize items in music datasets. These algorithms build summaries (both concise and diverse), by selecting appropriate segments from the input signal, also making them good candidates to summarize music. We evaluate the summarization process on binary and multiclass music genre classification tasks, by comparing the accuracy when using summarized datasets against the accuracy when using human-oriented summaries, continuous segments (the traditional method used for addressing the previously mentioned time constraints), and full songs of the original dataset. We show that GRASSHOPPER, LexRank, LSA, MMR, and a Support Sets-based centrality model improve classification performance when compared to selected baselines. We also show that summarized datasets lead to a classification performance whose difference is not statistically significant from using full songs. Furthermore, we make an argument stating the advantages of sharing summarized datasets for future MIR research. |
publishDate |
2016 |
dc.date.none.fl_str_mv |
2016-12-15T11:18:16Z 2016-01-01T00:00:00Z 2016 2019-04-10T10:07:41Z |
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/10071/12269 |
url |
http://hdl.handle.net/10071/12269 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
2329-9290 10.1109/TASLP.2016.2541299 |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.publisher.none.fl_str_mv |
IEEE |
publisher.none.fl_str_mv |
IEEE |
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_ |
1799134745523650560 |