Skip to main content
Top
Published in: Automatic Control and Computer Sciences 4/2023

01-08-2023

Environmental Sound Classification Based on Attention Feature Fusion and Improved Residual Network

Authors: Jinfang Zeng, Yuxing Liu, Mengjiao Wang, Xin Zhang

Published in: Automatic Control and Computer Sciences | Issue 4/2023

Login to get access

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

The classification of environmental sound is an important research area in artificial intelligence and its classification accuracy is greatly affected by feature extraction. However, most existing methods for feature set generation use simple feature fusion methods, which are ineffective for multi-classification purposes. To solve this problem and improve the neural network classification performance of existing training environmental sound classification (ESC) tasks, we first add the Gaussian error linear unit (GELU) activation function and gated linear units (GLU) to the residual network, which improves the network’s stability. Subsequently, this paper proposes a feature fusion method based on the attention mechanism and employs squeeze-and-excitation networks (SENet) to make network learning features fusion and training more successfully, which offers obvious advantages over existing classification methods. Experimental results show that our model has reached an obvious increase in classification accuracy for the two datasets i.e. ESC-10 (98.27%) and ESC-50 (98.32%).
Literature
4.
go back to reference Vacher, M., Istrate, D., Besacier, L., Serignat, J., and Castelli, E., Sound detection and classification for medical telesurvey, 2nd Conf. on Biomedical Engineering, Innsbruck, Austria, 2004, Calgary: ACTA Press, 2004, pp. 395–398. https://hal.science/hal-01088243. Vacher, M., Istrate, D., Besacier, L., Serignat, J., and Castelli, E., Sound detection and classification for medical telesurvey, 2nd Conf. on Biomedical Engineering, Innsbruck, Austria, 2004, Calgary: ACTA Press, 2004, pp. 395–398. https://​hal.​science/​hal-01088243.​
13.
go back to reference Dauphin, Ya.N., Fan, A., Auli, M., and Grangier, D., Language modeling with gated convolutional networks, Proc. Mach. Learn. Res., 2017, vol. 70, pp. 933–941. Dauphin, Ya.N., Fan, A., Auli, M., and Grangier, D., Language modeling with gated convolutional networks, Proc. Mach. Learn. Res., 2017, vol. 70, pp. 933–941.
19.
go back to reference Zhang, Z., Xu, S., Qiao, T., Zhang, S., and Cao, S., Attention based convolutional recurrent neural network for environmental sound classification, Pattern Recognition and Computer Vision, Lin, Z., Wang, L., Yang, J., Eds., Lecture Notes in Computer Science, vol. 11857, Cham: Springer, 2019, pp. 261–271. https://doi.org/10.1007/978-3-030-31654-9_23CrossRef Zhang, Z., Xu, S., Qiao, T., Zhang, S., and Cao, S., Attention based convolutional recurrent neural network for environmental sound classification, Pattern Recognition and Computer Vision, Lin, Z., Wang, L., Yang, J., Eds., Lecture Notes in Computer Science, vol. 11857, Cham: Springer, 2019, pp. 261–271. https://​doi.​org/​10.​1007/​978-3-030-31654-9_​23CrossRef
20.
go back to reference Bae, S., Choi, I., and Kim, N., Acoustic scene classification using parallel combination of LSTM and CNN, Detection Classification of Acoustic Scenes and Events, Budapest, 2016, pp. 11–15. Bae, S., Choi, I., and Kim, N., Acoustic scene classification using parallel combination of LSTM and CNN, Detection Classification of Acoustic Scenes and Events, Budapest, 2016, pp. 11–15.
26.
Metadata
Title
Environmental Sound Classification Based on Attention Feature Fusion and Improved Residual Network
Authors
Jinfang Zeng
Yuxing Liu
Mengjiao Wang
Xin Zhang
Publication date
01-08-2023
Publisher
Pleiades Publishing
Published in
Automatic Control and Computer Sciences / Issue 4/2023
Print ISSN: 0146-4116
Electronic ISSN: 1558-108X
DOI
https://doi.org/10.3103/S0146411623040119

Other articles of this Issue 4/2023

Automatic Control and Computer Sciences 4/2023 Go to the issue