Better Never Than Late: Timely Edge Video Analytics over the Air

Vinod Nigade; Ramon Winder; Henri Bal; Lin Wang

doi:10.1145/3485730.3493446

Better Never Than Late: Timely Edge Video Analytics over the Air

Vinod Nigade, Ramon Winder, Henri Bal, Lin Wang

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

Abstract

Edge video analytics based on deep learning has become an important building block for many modern intelligent applications such as mobile augmented reality and autonomous driving. Various mechanisms have been developed to handle dynamic wireless networks, compute resource availability, and achieve high analytics accuracy via filtering, DNN compression, pruning, and adaptation. So far, limited attention has been paid to timeliness - -providing strict service-level objectives (SLO) for edge video analytics pipelines, which is essential for the usability of user-interactive and mission-critical intelligent applications. In this paper, we analyze the challenges in achieving SLO for edge video analytics and present a system design for timely edge video analytics over the air leveraging a simple yet effective idea - -feedback control. Our preliminary evaluation based on a system prototype and real-world network traces shows the potential of our design. We also discuss the limitations, calling for future work.

Original language	English
Title of host publication	SenSys 2021
Subtitle of host publication	Proceedings of the 2021 19th ACM Conference on Embedded Networked Sensor Systems
Publisher	Association for Computing Machinery, Inc
Pages	426-432
Number of pages	7
ISBN (Electronic)	9781450390972
DOIs	https://doi.org/10.1145/3485730.3493446
Publication status	Published - Nov 2021
Event	19th ACM Conference on Embedded Networked Sensor Systems, SenSys 2021 - Coimbra, Portugal Duration: 15 Nov 2021 → 17 Nov 2021

Conference

Conference	19th ACM Conference on Embedded Networked Sensor Systems, SenSys 2021
Country/Territory	Portugal
City	Coimbra
Period	15/11/21 → 17/11/21

Bibliographical note

Funding Information:
This work is part of the Efficient Deep Learning (EDL) programme (grant number P16-25), financed by the Dutch Research Council (NWO). Lin Wang was partially supported by the German Research Foundation (DFG) Collaborative Research Center 1053–MAKI.

Publisher Copyright:
© 2021 Owner/Author.

Funding

This work is part of the Efficient Deep Learning (EDL) programme (grant number P16-25), financed by the Dutch Research Council (NWO). Lin Wang was partially supported by the German Research Foundation (DFG) Collaborative Research Center 1053–MAKI.

Keywords

edge computing
SLO guarantee
video analytics

Access to Document

10.1145/3485730.3493446

Persistent URL (handle)

Persistent link

Cite this

@inproceedings{ced715c19e594a3a8c37ce2153478055,

title = "Better Never Than Late: Timely Edge Video Analytics over the Air",

abstract = "Edge video analytics based on deep learning has become an important building block for many modern intelligent applications such as mobile augmented reality and autonomous driving. Various mechanisms have been developed to handle dynamic wireless networks, compute resource availability, and achieve high analytics accuracy via filtering, DNN compression, pruning, and adaptation. So far, limited attention has been paid to timeliness - -providing strict service-level objectives (SLO) for edge video analytics pipelines, which is essential for the usability of user-interactive and mission-critical intelligent applications. In this paper, we analyze the challenges in achieving SLO for edge video analytics and present a system design for timely edge video analytics over the air leveraging a simple yet effective idea - -feedback control. Our preliminary evaluation based on a system prototype and real-world network traces shows the potential of our design. We also discuss the limitations, calling for future work.",

keywords = "edge computing, SLO guarantee, video analytics",

author = "Vinod Nigade and Ramon Winder and Henri Bal and Lin Wang",

note = "Funding Information: This work is part of the Efficient Deep Learning (EDL) programme (grant number P16-25), financed by the Dutch Research Council (NWO). Lin Wang was partially supported by the German Research Foundation (DFG) Collaborative Research Center 1053–MAKI. Publisher Copyright: {\textcopyright} 2021 Owner/Author.; 19th ACM Conference on Embedded Networked Sensor Systems, SenSys 2021 ; Conference date: 15-11-2021 Through 17-11-2021",

year = "2021",

month = nov,

doi = "10.1145/3485730.3493446",

language = "English",

pages = "426--432",

booktitle = "SenSys 2021",

publisher = "Association for Computing Machinery, Inc",

}

Nigade, V, Winder, R, Bal, H & Wang, L 2021, Better Never Than Late: Timely Edge Video Analytics over the Air. in SenSys 2021: Proceedings of the 2021 19th ACM Conference on Embedded Networked Sensor Systems. Association for Computing Machinery, Inc, pp. 426-432, 19th ACM Conference on Embedded Networked Sensor Systems, SenSys 2021, Coimbra, Portugal, 15/11/21. https://doi.org/10.1145/3485730.3493446

Better Never Than Late: Timely Edge Video Analytics over the Air. / Nigade, Vinod; Winder, Ramon; Bal, Henri et al.
SenSys 2021: Proceedings of the 2021 19th ACM Conference on Embedded Networked Sensor Systems. Association for Computing Machinery, Inc, 2021. p. 426-432.

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Better Never Than Late

T2 - 19th ACM Conference on Embedded Networked Sensor Systems, SenSys 2021

AU - Nigade, Vinod

AU - Winder, Ramon

AU - Bal, Henri

AU - Wang, Lin

N1 - Funding Information: This work is part of the Efficient Deep Learning (EDL) programme (grant number P16-25), financed by the Dutch Research Council (NWO). Lin Wang was partially supported by the German Research Foundation (DFG) Collaborative Research Center 1053–MAKI. Publisher Copyright: © 2021 Owner/Author.

PY - 2021/11

Y1 - 2021/11

N2 - Edge video analytics based on deep learning has become an important building block for many modern intelligent applications such as mobile augmented reality and autonomous driving. Various mechanisms have been developed to handle dynamic wireless networks, compute resource availability, and achieve high analytics accuracy via filtering, DNN compression, pruning, and adaptation. So far, limited attention has been paid to timeliness - -providing strict service-level objectives (SLO) for edge video analytics pipelines, which is essential for the usability of user-interactive and mission-critical intelligent applications. In this paper, we analyze the challenges in achieving SLO for edge video analytics and present a system design for timely edge video analytics over the air leveraging a simple yet effective idea - -feedback control. Our preliminary evaluation based on a system prototype and real-world network traces shows the potential of our design. We also discuss the limitations, calling for future work.

AB - Edge video analytics based on deep learning has become an important building block for many modern intelligent applications such as mobile augmented reality and autonomous driving. Various mechanisms have been developed to handle dynamic wireless networks, compute resource availability, and achieve high analytics accuracy via filtering, DNN compression, pruning, and adaptation. So far, limited attention has been paid to timeliness - -providing strict service-level objectives (SLO) for edge video analytics pipelines, which is essential for the usability of user-interactive and mission-critical intelligent applications. In this paper, we analyze the challenges in achieving SLO for edge video analytics and present a system design for timely edge video analytics over the air leveraging a simple yet effective idea - -feedback control. Our preliminary evaluation based on a system prototype and real-world network traces shows the potential of our design. We also discuss the limitations, calling for future work.

KW - edge computing

KW - SLO guarantee

KW - video analytics

UR - http://www.scopus.com/inward/record.url?scp=85120913142&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85120913142&partnerID=8YFLogxK

U2 - 10.1145/3485730.3493446

DO - 10.1145/3485730.3493446

M3 - Conference contribution

AN - SCOPUS:85120913142

SP - 426

EP - 432

BT - SenSys 2021

PB - Association for Computing Machinery, Inc

Y2 - 15 November 2021 through 17 November 2021

ER -

Better Never Than Late: Timely Edge Video Analytics over the Air

Abstract

Conference

Bibliographical note

Funding

Keywords

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this