Towards Real-Time CNN Inference from a Video Stream on a Mobile GPU (WiP Paper)

Chanyoung Oh, Gunju Park, Sumin Kim, Dohee Kim, Youngmin Yi

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

5 Scopus citations

Abstract

While there are several frameworks for CNN inference on mobile GPUs, they do not achieve real-Time processing for the most of the CNNs that aim at reasonable accuracy since they all employ kernel-by-kernel execution model and do not effectively support INT8 quantization yet. In this paper, we reveal that mobile GPUs suffer from large kernel launch overhead unlike server GPUs, and then propose an on-device deep learning inference framework that can achieve real-Time inference of CNNs on mobile GPUs by removing kernel launch overhead and by effectively exploiting INT8 quantization. We have evaluated the proposed framework with a state-of-The-Art CNN based face detector (RetinaFace), and observed up to 2.01X of speedup compared to ARM Compute Library (ACL) on a commodity smartphone.

Original languageEnglish
Title of host publicationLCTES 2020 - 21st ACM SIGPLAN/SIGBED Conference on Languages, Compilers, and Tools for Embedded Systems
PublisherAssociation for Computing Machinery
Pages136-140
Number of pages5
ISBN (Electronic)9781450370943
DOIs
StatePublished - 16 Jun 2020
Event21st ACM SIGPLAN/SIGBED Conference on Languages, Compilers, and Tools for Embedded Systems, LCTES 2020 - London, United Kingdom
Duration: 16 Jun 2020 → …

Publication series

NameProceedings of the ACM SIGPLAN Conference on Languages, Compilers, and Tools for Embedded Systems (LCTES)

Conference

Conference21st ACM SIGPLAN/SIGBED Conference on Languages, Compilers, and Tools for Embedded Systems, LCTES 2020
Country/TerritoryUnited Kingdom
CityLondon
Period16/06/20 → …

Keywords

  • face detection
  • on-device deep learning
  • persistent threads
  • quantization

Fingerprint

Dive into the research topics of 'Towards Real-Time CNN Inference from a Video Stream on a Mobile GPU (WiP Paper)'. Together they form a unique fingerprint.

Cite this