EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Joint Learning of Emotion and Singing Style for Enhanced Music Style Understanding

Yuwen Chen, Jing Mao, Rui-Feng Wang iD

Sensors · 2025

DOI: 10.3390/s25247575 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, 2025. Verfasst von Yuwen Chen und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Humanities and Arts, Hunan Institute of Traffic Engineering, Hengyang 421219, China. Hauptthemen laut MEDLINE: Emotions, Music, Singing, Deep Learning. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hunan Provincial Education Science Planning Project, 2025 General Teaching Reform Project of Ordinary Undergraduate Universities in Hunan Province.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Understanding music styles is essential for music information retrieval, personalized recommendation, and AI-assisted content creation. However, existing work typically addresses tasks such as emotion classification and singing style classification independently, thereby neglecting the intrinsic relationships between them. In this study, we introduce a multi-task learning framework that jointly models these two tasks to enable explicit knowledge sharing and mutual enhancement. Our results indicate that joint optimization consistently outperforms single-task counterparts, demonstrating the value of leveraging inter-task correlations for more robust singing style analysis. To assess the generality and adaptability of the proposed framework, we evaluate it across various backbone architectures, including Transformer, TextCNN, and BERT, and observe stable performance improvements in all cases. Experiments on a benchmark dataset, which were self-constructed and collected through professional recording devices, further show that the framework not only achieves the best accuracy on both tasks on our dataset under a singer-wise split, but also yields interpretable insights into the interplay between emotional expression and stylistic characteristics in vocal performance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

EmotionsMusicSingingDeep LearningHumansComprehensionConvolutional Neural NetworksDatasets as TopicInformation DisseminationLarge Language Models

Stichwörter der Autor:innen

AI4Musicmulti-task learningmusic emotionmusic singing style

Autor:innen und Institutionen

  • School of Humanities and Arts, Hunan Institute of Traffic Engineering, Hengyang 421219, China.
  • Department of Crop and Soil Sciences, College of Agriculture and Environmental Sciences, University of Georgia, Tifton, GA 31793, USA.

ORCID-Profile: Rui-Feng Wang

Förderung

  • Hunan Provincial Education Science Planning Project
  • 2025 General Teaching Reform Project of Ordinary Undergraduate Universities in Hunan Province

Zitieren

APA

Chen, Y., Mao, J., & Wang, R. F. (2025). Joint Learning of Emotion and Singing Style for Enhanced Music Style Understanding. Sensors. https://doi.org/10.3390/s25247575

Vancouver

Chen Y, Mao J, Wang RF. Joint Learning of Emotion and Singing Style for Enhanced Music Style Understanding. Sensors. 2025. doi:10.3390/s25247575

BibTeX
@article{Chen2025,
  author = {Chen, Yuwen and Mao, Jing and Wang, Rui-Feng},
  title = {Joint Learning of Emotion and Singing Style for Enhanced Music Style Understanding},
  journal = {Sensors},
  year = {2025},
  doi = {10.3390/s25247575}
}
Kostenlos starten →