|
|
在MCU端部署GRU模型实现鼾声检测
|
Abstract:
本研究旨在开发一种在资源受限的微控制器单元(MCU)上运行的方法，用以进行鼾声检测。不同于使用CNN进行声音检测的方式，我们采用门控循环单元(GRU)模型以对音频数据进行处理和分析。通过采用优化模型结构、模型量化等常用的模型优化方式，我们最终成功将GRU模型适配到低功耗的MCU平台，使其能够在不依赖外部计算资源的情况下，独立完成端侧的鼾声检测任务，无需联网。实验结果表明，该模型在保持较高准确性的同时，能够有效降低系统算力需求，满足移动健康监测设备的实时性与便携性要求。这一研究为鼾症患者的持续监测和睡眠健康管理提供了一种新的解决方案，同时也拓展了深度学习在嵌入式系统中的应用前景。
This study aims to develop a method running on a resource-constrained microcontroller unit (MCU) for snore detection. Unlike the approach of using CNNs for sound detection, we employ the Gated Recurrent Unit (GRU) model to process and analyze audio data. By adopting common model optimization techniques such as optimizing the model structure and model quantization, we ultimately succeeded in adapting the GRU model to a low-power MCU platform. This allows it to independently perform snore detection tasks on the edge without relying on external computing resources and without the need for internet connectivity. Experimental results indicate that while maintaining high accuracy, the model can effectively reduce the system’s computational power requirements, meeting the real-time and portability needs of mobile health monitoring devices. This research provides a new solution for the continuous monitoring and sleep health management of patients with snoring disorders and also expands the application prospects of deep learning in embedded systems.
| [1] | Nguyen, M. and Huang, J. (2022) Snore Detection Using Convolution Neural Networks and Data Augmentation. In: Long, B.T., Kim, H.S., Ishizaki, K., Toan, N.D., Parinov, I.A. and Kim, YH., Eds., Proceedings of the International Conference on Advanced Mechanical Engineering, Automation, and Sustainable Development 2021 (AMAS2021). AMAS 2021. Lecture Notes in Mechanical Engineering. Springer, Cham. https://doi.org/10.1007/978-3-030-99666-6_15 |
| [2] | Xie, J., Aubert, X., Long, X., van Dijk, J., Arsenali, B., Fonseca, P., et al. (2021) Audio-Based Snore Detection Using Deep Neural Networks. Computer Methods and Programs in Biomedicine, 200, Article 105917. https://doi.org/10.1016/j.cmpb.2020.105917 |
| [3] | Goodfellow, I., Bengio, Y. and Courville, A. (2016) Deep Learning. MIT Press, 367-415. |
| [4] | Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., et al. (2014). Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), Doha, October 2014, 1724-1734. https://doi.org/10.3115/v1/d14-1179 |
| [5] | Zhang, A., Lipton, Z.C., Li, M. and Smola, A.J. (2023) Dive into Deep Learning. Cambridge University Press. |
| [6] | Krishnamoorthi, R. (2018) Quantizing Deep Convolutional Networks for Efficient Inference: A Whitepaper. arXiv: 1806.08342. https://doi.org/10.48550/arXiv.1806.08342 |
