Conference paper Open Access
Xygkis, Athanasios; Papadopoulos, Lazaros; Moloney, David; Soudris, Dimitrios; Yous, Sofiane
<?xml version='1.0' encoding='utf-8'?> <oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd"> <dc:creator>Xygkis, Athanasios</dc:creator> <dc:creator>Papadopoulos, Lazaros</dc:creator> <dc:creator>Moloney, David</dc:creator> <dc:creator>Soudris, Dimitrios</dc:creator> <dc:creator>Yous, Sofiane</dc:creator> <dc:date>2018-09-20</dc:date> <dc:description>The implementation of Convolutional Neural Networks on edge Internet of Things (IoT) devices is a significant programming challenge, due to the limited computational resources and the real-time requirements of modern applications. This work focuses on the efficient implementation of the Winograd convolution, based on a set of application-independent and Winograd-specific software techniques for improving the utilization of the edge devices computational resources. The proposed techniques were evaluated in Intel/Movidius Myriad2 platform, using 4 CNNs of various computational requirements. The results show significant performance improvements, up to 54%, over other convolution algorithms.</dc:description> <dc:identifier>https://zenodo.org/record/3347180</dc:identifier> <dc:identifier>10.1145/3195970.3196041</dc:identifier> <dc:identifier>oai:zenodo.org:3347180</dc:identifier> <dc:relation>info:eu-repo/grantAgreement/EC/H2020/780572/</dc:relation> <dc:rights>info:eu-repo/semantics/openAccess</dc:rights> <dc:rights>https://creativecommons.org/licenses/by/4.0/legalcode</dc:rights> <dc:title>Efficient Winograd-based Convolution Kernel Implementation on Edge Devices</dc:title> <dc:type>info:eu-repo/semantics/conferencePaper</dc:type> <dc:type>publication-conferencepaper</dc:type> </oai_dc:dc>
Views | 53 |
Downloads | 1,854 |
Data volume | 650.1 MB |
Unique views | 47 |
Unique downloads | 1,786 |