Hardware Acceleration of Capsule Networks for Real-time Applications

被引：0

作者：

Hemmati, Maryam ^{[1
]}

Babette, Earlene Starling ^{[1
]}

Shan, Julia ^{[1
]}

Biglari-Abhari, Morteza ^{[1
]}

Niar, Smail ^{[2
,3
]}

机构：

[1] Univ Auckland, Dept Elect Comp & Software Engn, Auckland, New Zealand

[2] Univ Polytech Hauts de France, CNRS, LAMIH, Valenciennes, France

[3] Univ Polytech Hauts de France, INSA Hauts de France, Valenciennes, France

来源：

2024 27TH EUROMICRO CONFERENCE ON DIGITAL SYSTEM DESIGN, DSD 2024 | 2024年

关键词：

Capsule Network; Real-time Application; Hardware Accelerator; FPGA;

D O I：

10.1109/DSD64264.2024.00012

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

Capsule networks (CapsNet) are a category of deep learning neural networks (DNN) that address one of the main issues and deficiencies of convolutional neural networks (CNN); loss of spatial information in pooling layers. However, the main concern with CapsNets is their compute-intensive nature, which is mainly related to the vector-based calculations during dynamic routing and acts as a barrier for their deployment in real-time applications. To address the specific computing requirements of dynamic routing in capsule layers of CapsNets, we develop a hardware accelerator for dynamic routing using Vitis HLS. In this paper, we present a hardware acceleration solution for capsule networks by integrating AMD Xilinx deep processing unit (DPU) and a custom accelerator for the capsule layer. Our results show significant improvement in throughput compared with the baseline implementation when CapsNet is implemented on Zynq UltraScale+ MPSoC ZCU102 using Vitis AI DPUs.

引用

页码：19 / 25

页数：7

共 28 条

[21]

Shahroudnejad A, 2018, Arxiv, DOI arXiv:1802.10204

[22]

Szegedy C, 2014, Arxiv, DOI [arXiv:1312.6199, DOI 10.1109/CVPR.2015.7298594]

[23] Convolutional capsule network for COVID-19 detection using radiography images [J].