This paper studies the automatic recognition system of vocal performance style based on residual network (ResNet). The system can precisely recognize multiple vocal performance styles through automation technology. Firstly, an audio dataset covering multiple vocal styles was constructed based on the publicly available dataset GTZAN, and the data in the dataset was preprocessed to extract key music features. Subsequently, the ResNet model was chosen as the core recognition component of the system, which was trained to learn the complex relationship between audio features and vocal styles. Finally, the construction of the system was carried out. In the verification phase, the dataset was divided into two parts: the training set and the test set. After training the model using the training set, its performance was comprehensively evaluated on the test set. The experiment showed that the recognition system based on ResNet performed well in precision, recall, and accuracy, with specific values of 95.6%, 91.6%, and 96.7%, respectively. Meanwhile, the recognition accuracy of the ResNet-based system for 10 vocal performance styles was 95.96% on average. Applying ResNet to an automatic recognition system for vocal performance styles not only provides an efficient and reliable solution in this field, but also opens up new horizons for the further application of deep learning in the music industry.
TanDDiazFMMikszaP. Expressing emotion through vocal performance: acoustic cues and the effects of a mindfulness induction. Psychol Music2020; 48(4): 495–512. DOI: 10.1177/0305735618809873.
2.
BhatJPRajalakshmiK. Pitch coding in vocalists and non musicians to Carnatic music stimuli: a frequency following response (FFR) study. Int J Health Sci Res2020; 10(10): 16–21. DOI: 10.20517/ss.2022.06.
3.
BeibeiW. Automatic integration algorithm of vocal performance learning materials based on multidimensional association rules. Math Probl Eng2022; 2022(1): 6863141–6863149. DOI: 10.1155/2022/6863141.
4.
HakanpaaTWaaramaaTLaukkanenA-M. Emotion recognition from singing voices using contemporary commercial music and classical styles. J Voice2019; 33(4): 501–509. DOI: 10.1016/j.jvoice.2018.01.012.
5.
ZhangXYuYGaoY, et al.Research on singing voice detection based on a long-term recurrent convolutional network with vocal separation and temporal smoothing. Electronics2020; 9(9): 1458. DOI: 10.1155/2022/5111896.
6.
WadiyoWHaryonoSWiyosoJ, et al.Aesthetic features and ethnic music style in social and cultural life of the modern era. Int Rev Aesthet Sociol Music2021; 52(1): 83–100. https://www.jstor.org/stable/27032559
7.
FalolaPBAkinolaSO. Music genre classification using 1D convolution neural network. Int J Hum Comput Stud2021; 3(6): 3–21. https://www.jstor.org/stable/27032559
8.
SiyuJJingyanQ. Research on music visualization intelligent design based on computer image style transfer. Packaging Engineering Art Edition2020; 41(16): 193–198. DOI: 10.19554/j.cnki.1001-3563.2020.16.029.
9.
SolankiAPandeyS. Music instrument recognition using deep convolutional neural networks. Int J Inf Technol2022; 14(3): 1659–1668. DOI: 10.1007/s41870-019-00285-y.
10.
YanF. Music recognition algorithm based on TS cognitive neural network. Transl Neurosci2019; 10(1): 135–140. DOI: 10.1515/tnsci-2019-0023.
11.
MaqboolKSinghRPMehraM. A CNN and ResNet 50 approach to detect cardiac diseases using ECG images. Int J Res Appl Sci Eng Technol2022; 10(11): 641–651.
12.
YingCXiangWHuangZ. Underdetermined DOA estimation via multiple time-delay covariance matrices and deep residual network. J Syst Eng Electron2021; 32(6): 1354–1363. DOI: 10.23919/JSEE.2021.000115.
13.
LanRSunLLiuZ, et al.Cascading and enhanced residual networks for accurate single-image super-resolution. IEEE Trans Cybern2020; 51(1): 115–125. DOI: 10.1109/TCYB.2019.2952710.
14.
HanXChenWZhouC. Musical genre classification based on deep residual auto-encoder and support vector machine. Journal of Information Processing Systems2024; 20(1): 13–23. DOI: 10.3745/JIPS.04.0300.
15.
GauerJNagathilAEckelK, et al.A versatile deep-neural-network-based music preprocessing and remixing scheme for cochlear implant listeners. J Acoust Soc Am2022; 151(5): 2975–2986. DOI: 10.1121/10.0010371.
16.
HanXChenFBanJ. Music emotion recognition based on a neural network with an Inception-GRU residual structure. Electronics2023; 12(4): 978. DOI: 10.3390/electronics12040978.
17.
ZhaoXWangYCaiX. A ResNet-based audio-visual fusion model for piano skill evaluation. Appl Sci2023; 13(13): 7431–7436. DOI: 10.3390/app13137431.
18.
MirzaFKGursoyAFBaykasT, et al.Residual LSTM neural network for time dependent consecutive pitch string recognition from spectrograms: a study on Turkish classical music makams. Multimed Tool Appl2024; 83(14): 41243–41271. DOI: 10.1007/s11042-023-17105-y.
19.
AbdallahSEElmesseryWMShamsMY, et al.Deep learning model based on ResNet-50 for beef quality classification. Inf Sci Lett2023; 12(1): 289–297.
20.
WanjunLRenjieMHaichengQ, et al.Research on music genre recognition based on enhanced AlexNet. CAAI Transactions on Intelligent Systems2020; 15(4): 750–757. DOI: 10.11992/tis.201909032.
21.
JiaoJMaSSongY, et al.MGTN, a bimodal music genre classification model based on convolutional attention mechanism. Computer Engineering and Science2023; 45(12): 2226. https://joces.nudt.edu.cn/CN/Y2023/V45/I12/2226
22.
KhaliliGAAliKKYilmazerR, et al.Local fractal Fourier transform and applications. Computational Methods for Differential Equations2022; 10(3): 595–607. DOI: 10.22034/cmde.2021.42554.1832.
23.
WangZBaladron-ZoritaOHellmannC, et al.Theory and algorithm of the homeomorphic Fourier transform for optical simulations. Opt Express2020; 28(7): 10552–10571. DOI: 10.1364/OE.388022.
24.
NicolaFTilliP. The Faber–Krahn inequality for the short-time Fourier transform. Invent Math2022; 230(1): 1–30. DOI: 10.37965/jdmd.v2i2.39.
25.
PrzystupaKAmbrozkiewiczBLitakG. Diagnostics of transient states in hydraulic pump system with Short Time Fourier Transform. Adv Sci Technol Res J2020; 14(1): 178–183. DOI: 10.12913/22998624/116971.
26.
LokeshSRamya DeviM. Speech recognition system using enhanced mel frequency cepstral coefficient with windowing and framing method. Cluster Comput2019; 22(Suppl 5): 11669–11679. DOI: 10.1007/s10586-017-1447-6.
27.
NguyenMTLinWWHuangJH. Heart sound classification using deep learning techniques based on log-mel spectrogram. Circ Syst Signal Process2023; 42(1): 344–360. DOI: 10.1007/s00034-022-02124-1.
28.
KarimAASalehSM. Text to speech using Mel-Spectrogram with deep learning algorithms. Period Eng Nat Sci2022; 10(3): 380–386. DOI: 10.21533/pen.v10i3.3113.
29.
YangNDeyNSherrattRS, et al.Recognize basic emotional statesin speech by machine learning techniques using mel-frequency cepstral coefficient features. J Intell Fuzzy Syst2020; 39(2): 1925–1936. DOI: 10.3233/JIFS-179963.
30.
WangRCAnSLiJ, et al.Deep residual network framework for structural health monitoring. Struct Health Monit2021; 20(4): 1443–1461. DOI: 10.1177/1475921720918378.
31.
ZhaoMZhongSFuX, et al.Deep residual networks with adaptively parametric rectifier linear units for fault diagnosis. IEEE Trans Ind Electron2020; 68(3): 2587–2597. DOI: 10.1109/TIE.2020.2972458.
32.
ZuoYWuQFangY, et al.Multi-scale frequency reconstruction for guided depth map super-resolution via deep residual network. IEEE Trans Circ Syst Video Technol2019; 30(2): 297–306. DOI: 10.1109/TCSVT.2018.2890271.
33.
LiuWWuGRenF, et al.DFF-ResNet: an insect pest recognition model based on residual networks. Big Data Min Anal2020; 3(4): 300–310. DOI: 10.26599/BDMA.2020.9020021.
34.
AliYMNoorsalEMokhtarNF, et al.Speech-based gender recognition using linear prediction and mel-frequency cepstral coefficients. Indonesian J Electr Eng Comput Sci2022; 28(2): 753–761. DOI: 10.1007/978-3-031-48309-7_10.
35.
ChenKLiuZLiuQ, et al.EEG-based mental fatigue detection using linear prediction cepstral coefficients and Riemann spatial covariance matrix. J Neural Eng2022; 19(6): 066021. DOI: 10.1088/1741-2552/aca1e2.
36.
ChenSAhmmedSLalK, et al.Django web development framework: powering the modern web. Am J Trade Policy2020; 7(3): 99–106. DOI: 10.18034/ajtp.v7i3.675.
