Affiliation:
1. Department of Electrical and Information Engineering, Beijing University of Civil Engineering and Architecture, Beijing 100044, China
Abstract
Music emotion recognition has garnered significant interest in recent years, as the emotions expressed through music can profoundly enhance our understanding of its deeper meanings. The violin, with its distinctive emotional expressiveness, has become a focal point in this field of research. To address the scarcity of specialized data, we developed a dataset specifically for violin music emotion recognition named VioMusic. This dataset offers a precise and comprehensive platform for the analysis of emotional expressions in violin music, featuring specialized samples and evaluations. Moreover, we implemented the CNN–BiGRU–Attention (CBA) model to establish a baseline system for music emotion recognition. Our experimental findings show that the CBA model effectively captures the emotional nuances in violin music, achieving mean absolute errors (MAE) of 0.124 and 0.129. The VioMusic dataset proves to be highly practical for advancing the study of emotion recognition in violin music, providing valuable insights and a robust framework for future research.
Reference32 articles.
1. Sun, H. (2023). China’s Music Industry to Total over 378.7 Billion Yuan by 2021, China Press, Publication, Radio and Television News.
2. Speech emotion recognition using hybrid spectral-prosodic features of speech signal/glottal waveform, metaheuristic-based dimensionality reduction, and Gaussian elliptical basis function network classifier;Daneshfar;Appl. Acoust.,2020
3. Matsunaga, M., Kikusui, T., Mogi, K., Nagasawa, M., and Myowa, M. (2020). Breastfeeding dynamically changes endogenous oxytocin levels and emotion recognition in mothers. Biol. Lett., 16.
4. Improving cross-corpus speech emotion recognition using deep local domain adaptation;Zhao;Chin. J. Electron.,2022
5. Yang, P.T., Kuang, S.M., Wu, C.C., and Hsu, J.L. (2020, January 19–24). Predicting music emotion by using convolutional neural network. Proceedings of the 22nd HCI International Conference, Copenhagen, Denmark.