基于硬件加速的轻量级网络心音分类器

doi:10.3778/j.issn.1002-8331.2103-0006

摘要/Abstract

摘要：

近年来，卷积神经网络被广泛应用于心音信号分类。为满足先心病机器辅助诊断系统低功耗、可移动等方面需求，基于轻量级神经网络MobileNet，实现了一种适用于FPGA硬件平台的心音分类器。心音分类器的深度卷积、逐点卷积与最大池化等模块通过高层次综合进行设计。该心音分类器在利用深度可分离卷积减少网络参数与运算量的同时，通过多像素多通道并行及定点量化等方式，提升了分类器运行速度。经心音数据集实验结果表明，在计算效率方面，该心音分类器在FPGA上相较于在通用CPU上实现约14倍加速。

关键词: 硬件加速, FPGA, MobileNet, 心音信号, 高层次综合

Abstract:

In recent years, convolutional neural networks have been widely used in heart sound signal classification. In order to meet the requirements of low power consumption and mobility of the machine-assisted diagnosis system for congenital heart disease, a kind of heart sound classifier suitable for FPGA hardware platform based on the lightweight neural network MobileNet is put forward in this paper. The deep convolution, point-by-point convolution, and maximum pooling modules of the heart sound classifier are designed through HLS. To reduce network parameters and calculations for the heart sound classifier, the deep separable convolution is used. At the same time, the running speed of the classifier is improved by using multi-pixel, multi-channel parallelism, and fixed-point quantization. The experimental results show that in terms of computational efficiency, compared with the traditional network, the heart sound classifier achieves approximately 14 times faster than general-purpose CPU platforms.

Key words: hardware acceleration, FPGA, MobileNet, heart sound signal, high level synthesis

杜煜章，潘家华，宗容，粟炜，王威廉. 基于硬件加速的轻量级网络心音分类器[J]. 计算机工程与应用, 2021, 57(23): 263-269.

DU Yuzhang, PAN Jiahua, ZONG Rong, SU Wei, WANG Weilian. Lightweight Network Heart Sound Classifier Based on Hardware Acceleration[J]. Computer Engineering and Applications, 2021, 57(23): 263-269.

参考文献

[1] 苏翠敏，林利平，蔡红珠，等.心脏杂音和POX在先心病筛查中的作用[J].中国医学创新，2019，16（30）：147-150.
SU Cuimin，LIN Liping，CAI Hongzhu，et al.The role of heart murmur and pulse oximetry in the screening of congenital heart disease[J].Medical Innovation of China，2019，16（30）：147-150.
[2] 谭朝文，王威廉，宗容，等.卷积神经网络应用于先心病心音信号分类研究[J].计算机工程与应用，2019，55（12）：174-180.
TAN Zhaowen，WANG Weilian，ZONG Rong，et al.Research on classification of congenital heart disease heart soundsignal using convolutional neural network[J].Computer Engineering and Applications，2019，55（12）：174-180.
[3] 陈伟，孙强，齐月月，等.深度卷积神经网络在心音分类方法中的应用[J].计算机工程与应用，2021，57（16）：182-189.
CHEN Wei，SUN Qiang，QI Yueyue，et al.Deep convolutional neural networks for heart sound classification[J].Computer Engineering and Applications，2021，57（16）：182-189.
[4] MAKNICKAS V，MAKNICKAS A.Recognition of normal-abnormal phonocardiographic signals using deep convolutional neural networks and mel-frequency spectral coefficients[J].Physiological Measurement，2017，38（8）：1671-1684.
[5] 陆志坚.基于FPGA的卷积神经网络并行结构研究[D].哈尔滨：哈尔滨工程大学，2013.
LU Zhijian.The research on parallel architecture for FPGA-based convolutional neural networks[D].Harbin：Harbin Engineering University，2013.
[6] 方睿，刘加贺，薛志辉，等.卷积神经网络的FPGA并行加速方案设计[J].计算机工程与应用，2015，51（8）：32-36.
FANG Rui，LIU Jiahe，XUE Zhihui，et al.FPGA-based design for convolution neural network[J].Computer Engineering and Applications，2015，51（8）：32-36.
[7] 陈辰，柴志雷，夏珺.基于Zynq7000 FPGA异构平台的YOLOv2加速器设计与实现[J].计算机科学与探索，2019，13（10）：1677-1693.
CHEN Chen，CHAI Zhilei，XIA Jun.Design and implementation of YOLOv2 accelerator based on Zynq7000 FPGA heterogeneous platform[J].Journal of Frontiers of Computer Science and Technology，2019，13（10）：1677-1693.
[8] 黄瑞，金光浩，李磊，等.轻量化神经网络加速器的设计与实现[J/OL].计算机工程：1-7.[2021-01-26].https：//doi.org/10.19678/j.issn.1000-3428.0059351.
HUANG Rui，JIN Guanghao，LI Lei，et al.Design and implementation of AI processor for lightweight neural network[J/OL].Computer Engineering：1-7[2021-01-26].https：//doi.org/10.19678/j.issn.1000-3428.0059351.
[9] 李炳辰，黄鲁.一种移动卷积神经网络的FPGA实现[J].微电子学与计算机，2019，36（9）：7-11.
LI Bingchen，HUANG Lu.Hardware implementation of a convolutional neural networkfor mobile terminal based on FPGA[J].Microelectronics & Computer，2019，36（9）：7-11.
[10] 陈辰.基于FPGA的神经网络加速器的规模可伸缩性研究[D].无锡：江南大学，2019.
CHEN Chen.Research of scalability on FPGA-based neural network accelerator[D].Wuxi：Jiangnan University，2019.
[11] HAN S，MAO H，DALLY W J.Deep compression：compressing deep neural networks with pruning，trained quantization and Huffman coding[J].Fiber，2015，56：3-7.
[12] ZHAO Y，GAO X，GUO X，et al.Automatic generation of multi-precision multi-arithmetic CNN accelerators for FPGAs[C]//2019 International Conference on Field-Programmable Technology（ICFPT），2019.
[13] HOWARD A G，ZHU M，CHEN B，et al.MobileNets：efficient convolutional neural networks for mobile vision applications[J].arXiv：1704.04861，2017.
[14] ZHANG C，LI P，SUN G Y，et al.Optimizing FPGA-based accelerator design for deep convolutional neural networks[C]//Proceedings of the 2015 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays，Monterey，Feb 22-24，2015.New York：ACM，2015：161-170.
[15] VENIERIS S I，BOUGANIS C S.fpgaConvNet： a framework for mapping convolutional neural networks on FPGAs[C]//IEEE International Symposium on Field-Programmable Custom Computing Machines，2016.
[16] 张卫，刘宇红，张荣芬.可实现时分复用的CNN卷积层和池化层IP核设计[J].计算机工程与应用，2020，56（24）：66-71.
ZHANG Wei，LIU Yuhong，ZHANG Rongfen.Design of IP cores for CNN convolutional layer and pooling layer capable of time division multiplexing[J].Computer Engineering and Applications，2020，56（24）：66-71.