COUPLE: Accelerating Video Analytics on Heterogeneous Mobile Processors

Hao Bao, Zhi Zhou, Jiajie Xie, Qianyi Huang, Fei Xu, Xu Chen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Deep learning has achieved tremendous success in various fields, but its significant computational demands make inference on mobile devices extremely challenging. To address this issue, we propose the COUPLE system, which enables heterogeneous processors to collaborate on mobile devices for accelerating video analytics. Additionally, we design the Co-Optimize strategy which utilizes the inference results of GPU to mitigate the accuracy loss caused by DSP. Experimental results demonstrate that COUPLE can improve the inference Average Precision by up to 5% compared to existing solutions.

Original languageEnglish
Title of host publicationProceedings of the 29th Annual International Conference on Mobile Computing and Networking, ACM MobiCom 2023
PublisherAssociation for Computing Machinery
Pages1552-1554
Number of pages3
ISBN (Electronic)9781450399906
DOIs
StatePublished - 2 Oct 2023
Event29th Annual International Conference on Mobile Computing and Networking, MobiCom 2023 - Madrid, Spain
Duration: 2 Oct 20236 Oct 2023

Publication series

NameProceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM
ISSN (Print)1543-5679

Conference

Conference29th Annual International Conference on Mobile Computing and Networking, MobiCom 2023
Country/TerritorySpain
CityMadrid
Period2/10/236/10/23

Keywords

  • deep learning
  • heterogeneous processors
  • mobile devices

Fingerprint

Dive into the research topics of 'COUPLE: Accelerating Video Analytics on Heterogeneous Mobile Processors'. Together they form a unique fingerprint.

Cite this