Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms

Jong Hwan Ko; Taesik Na; Mohammad Faisal Amir; Saibal Mukhopadhyay

doi:10.1109/AVSS.2018.8639121

Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms

Jong Hwan Ko
, Taesik Na
, Mohammad Faisal Amir
, Saibal Mukhopadhyay

Georgia Institute of Technology

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

146 Scopus citations

Abstract

This paper introduces partitioning an inference task of a deep neural network between an edge and a host platform in the IoT environment. We present a DNN as an encoding pipeline, and propose to transmit the output feature space of an intermediate layer to the host. Encoding of the feature space is proposed to enhance the maximum input rate supported by the edge platform and/or reduce the energy of the edge platform. Simulation results show that partitioning a DNN coupled with feature space encoding enables significant improvement in the energy-efficiency and throughput over the baseline configurations that perform the entire inference at the edge or at the host.

Original language	English
Title of host publication	Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781538692943
DOIs	https://doi.org/10.1109/AVSS.2018.8639121
State	Published - 2 Jul 2018
Externally published	Yes
Event	15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2018 - Auckland, New Zealand Duration: 27 Nov 2018 → 30 Nov 2018

Publication series

Name	Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance

Conference

Conference	15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2018
Country/Territory	New Zealand
City	Auckland
Period	27/11/18 → 30/11/18

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Access to Document

10.1109/AVSS.2018.8639121

Cite this

Ko, J. H., Na, T., Amir, M. F., & Mukhopadhyay, S. (2018). Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms. In Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance Article 8639121 (Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/AVSS.2018.8639121

Ko, Jong Hwan ; Na, Taesik ; Amir, Mohammad Faisal et al. / Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms. Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance. Institute of Electrical and Electronics Engineers Inc., 2018. (Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance).

@inproceedings{5a8fdfddb703477aa6a6fdcb5079e83a,

title = "Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms",

abstract = "This paper introduces partitioning an inference task of a deep neural network between an edge and a host platform in the IoT environment. We present a DNN as an encoding pipeline, and propose to transmit the output feature space of an intermediate layer to the host. Encoding of the feature space is proposed to enhance the maximum input rate supported by the edge platform and/or reduce the energy of the edge platform. Simulation results show that partitioning a DNN coupled with feature space encoding enables significant improvement in the energy-efficiency and throughput over the baseline configurations that perform the entire inference at the edge or at the host.",

author = "Ko, \{Jong Hwan\} and Taesik Na and Amir, \{Mohammad Faisal\} and Saibal Mukhopadhyay",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2018 ; Conference date: 27-11-2018 Through 30-11-2018",

year = "2018",

month = jul,

day = "2",

doi = "10.1109/AVSS.2018.8639121",

language = "English",

series = "Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance",

}

Ko, JH, Na, T, Amir, MF & Mukhopadhyay, S 2018, Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms. in Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance., 8639121, Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, Institute of Electrical and Electronics Engineers Inc., 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2018, Auckland, New Zealand, 27/11/18. https://doi.org/10.1109/AVSS.2018.8639121

Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms. / Ko, Jong Hwan; Na, Taesik; Amir, Mohammad Faisal et al.
Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance. Institute of Electrical and Electronics Engineers Inc., 2018. 8639121 (Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance).

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

TY - GEN

T1 - Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms

AU - Ko, Jong Hwan

AU - Na, Taesik

AU - Amir, Mohammad Faisal

AU - Mukhopadhyay, Saibal

PY - 2018/7/2

Y1 - 2018/7/2

N2 - This paper introduces partitioning an inference task of a deep neural network between an edge and a host platform in the IoT environment. We present a DNN as an encoding pipeline, and propose to transmit the output feature space of an intermediate layer to the host. Encoding of the feature space is proposed to enhance the maximum input rate supported by the edge platform and/or reduce the energy of the edge platform. Simulation results show that partitioning a DNN coupled with feature space encoding enables significant improvement in the energy-efficiency and throughput over the baseline configurations that perform the entire inference at the edge or at the host.

AB - This paper introduces partitioning an inference task of a deep neural network between an edge and a host platform in the IoT environment. We present a DNN as an encoding pipeline, and propose to transmit the output feature space of an intermediate layer to the host. Encoding of the feature space is proposed to enhance the maximum input rate supported by the edge platform and/or reduce the energy of the edge platform. Simulation results show that partitioning a DNN coupled with feature space encoding enables significant improvement in the energy-efficiency and throughput over the baseline configurations that perform the entire inference at the edge or at the host.

UR - https://www.scopus.com/pages/publications/85063273022

U2 - 10.1109/AVSS.2018.8639121

DO - 10.1109/AVSS.2018.8639121

M3 - Conference contribution

AN - SCOPUS:85063273022

T3 - Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance

BT - Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2018

Y2 - 27 November 2018 through 30 November 2018

ER -

Ko JH, Na T, Amir MF, Mukhopadhyay S. Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms. In Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance. Institute of Electrical and Electronics Engineers Inc. 2018. 8639121. (Proceedings of AVSS 2018 - 2018 15th IEEE International Conference on Advanced Video and Signal-Based Surveillance). doi: 10.1109/AVSS.2018.8639121

Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms

Abstract

Publication series

Conference

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this