EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A music structure analysis method based on beat feature and improved residual networks

Bing Lu, Qianxue Zhang, Yi Guo iD, Fuqiang Hu, Xuejun Xiong iD

PLOS ONE · 19.02.2025 · 20 (2) · e0312608

DOI: 10.1371/journal.pone.0312608 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, online erschienen am 19.02.2025. Verfasst von Bing Lu und 4 weiteren Autor:innen. Beteiligte Einrichtung: Xihua University, Chengdu, China. Hauptthemen laut MEDLINE: Music. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: The 2024 Xihua University Graduate Student Innovation and Entrepreneurship Competition Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In response to the issues of insufficient audio feature representation and insufficient model generalization ability in music structure analysis methods, a music structure analysis method based on beat feature fusion and an improved residual network was designed. Music structure analysis (MSA) contains two tasks, boundary detection and segment labeling. Boundary detection and Segment labeling are expected to accurately divide music segments and clarify the function. In this paper, a method is studied to accomplish the two tasks. First, the labels of music structure are refactored into 9 types and refined to each beat. Secondly, a beat-wise feature extraction is studied to segment music according to its beats and incorporates various acoustic features to achieve highly accurate segmentation. Then, a Resnet-34 based on a self-attentive mechanism is used to predict the category of each beat. Finally, a post-processing step is designed to filter predicted labels. The method is evaluated on the dataset SALAMI-IA, and experiments show that it is 3 percentage points higher than the current optimal method on HR3F. It is better than the most advanced methods on PWF and Sf.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

MusicHumansAlgorithmsAcousticsNeural Networks, Computer

Autor:innen und Institutionen

  • Xihua University, Chengdu, China.
  • Mashang Consumer Finance Co, Chongqing, China.

ORCID-Profile: Yi Guo, Xuejun Xiong

Förderung

  • The 2024 Xihua University Graduate Student Innovation and Entrepreneurship Competition Project

Zitieren

APA

Lu, B., Zhang, Q., Guo, Y., Hu, F., & Xiong, X. (2025). A music structure analysis method based on beat feature and improved residual networks. PLOS ONE, 20(2), e0312608. https://doi.org/10.1371/journal.pone.0312608

Vancouver

Lu B, Zhang Q, Guo Y, Hu F, Xiong X. A music structure analysis method based on beat feature and improved residual networks. PLOS ONE. 2025;20(2):e0312608. doi:10.1371/journal.pone.0312608

BibTeX
@article{Lu2025,
  author = {Lu, Bing and Zhang, Qianxue and Guo, Yi and Hu, Fuqiang and Xiong, Xuejun},
  title = {A music structure analysis method based on beat feature and improved residual networks},
  journal = {PLOS ONE},
  year = {2025},
  volume = {20},
  number = {2},
  pages = {e0312608},
  doi = {10.1371/journal.pone.0312608}
}
Kostenlos starten →