Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation

Asiri Lindamulage; Nuwan Kodagoda; Shyam Reyal; Pradeepa Samarasinghe; Pratheepan Yogarajah

doi:10.1109/tencon55691.2022.9977637

Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation

Asiri Lindamulage, Nuwan Kodagoda, Shyam Reyal, Pradeepa Samarasinghe, Pratheepan Yogarajah

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

Abstract

Magnitude-based pruning is a technique used to optimise deep learning models for edge inference. We have achieved over 75% model size reduction with a higher accuracy than the original multi-output regression model for head-pose estimation.

Original language	English
Title of host publication	Proceedings of TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON)
Publisher	IEEE
ISBN (Electronic)	978-1-6654-5095-9
ISBN (Print)	978-1-6654-5096-6
DOIs	https://doi.org/10.1109/tencon55691.2022.9977637
Publication status	Published (in print/issue) - 20 Dec 2022

Publication series

Name
Publisher	IEEE Control Society
ISSN (Print)	2159-3442
ISSN (Electronic)	2159-3450

Bibliographical note

Funding Information:
VI. EXPERIMENTAL RESULTS AND CODES Experimental results and the model pruning code can be found at: https://github.com/asirigawesha/PruningTests.git VII. ACKNOWLEDGEMENT This research was supported by the Accelerating Higher Education Expansion and Development (AHEAD) Operation of the Ministry of Higher Education of Sri Lanka funded by the World Bank (https://ahead.lk/result-area-3/).

Publisher Copyright:
© 2022 IEEE.

Keywords

Edge inference
Optimisation
Network Pruning
Quantisation
TensorFlow
Head Post estimation
Head Pose estimation
Edge Inference

Access to Document

10.1109/tencon55691.2022.9977637

Cite this

@inproceedings{3f54adb9bf6944489c1a35cf1c754e40,

title = "Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation",

abstract = "Magnitude-based pruning is a technique used to optimise deep learning models for edge inference. We have achieved over 75% model size reduction with a higher accuracy than the original multi-output regression model for head-pose estimation.",

keywords = "Edge inference, Optimisation, Network Pruning, Quantisation, TensorFlow, Head Post estimation, Head Pose estimation, Edge Inference",

author = "Asiri Lindamulage and Nuwan Kodagoda and Shyam Reyal and Pradeepa Samarasinghe and Pratheepan Yogarajah",

note = "Funding Information: VI. EXPERIMENTAL RESULTS AND CODES Experimental results and the model pruning code can be found at: https://github.com/asirigawesha/PruningTests.git VII. ACKNOWLEDGEMENT This research was supported by the Accelerating Higher Education Expansion and Development (AHEAD) Operation of the Ministry of Higher Education of Sri Lanka funded by the World Bank (https://ahead.lk/result-area-3/). Publisher Copyright: {\textcopyright} 2022 IEEE.",

year = "2022",

month = dec,

day = "20",

doi = "10.1109/tencon55691.2022.9977637",

language = "English",

isbn = "978-1-6654-5096-6",

publisher = "IEEE",

booktitle = "Proceedings of TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON)",

address = "United States",

}

Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation. / Lindamulage, Asiri; Kodagoda, Nuwan; Reyal, Shyam et al.
Proceedings of TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON). IEEE, 2022.

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

TY - GEN

T1 - Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation

AU - Lindamulage, Asiri

AU - Kodagoda, Nuwan

AU - Reyal, Shyam

AU - Samarasinghe, Pradeepa

AU - Yogarajah, Pratheepan

N1 - Funding Information: VI. EXPERIMENTAL RESULTS AND CODES Experimental results and the model pruning code can be found at: https://github.com/asirigawesha/PruningTests.git VII. ACKNOWLEDGEMENT This research was supported by the Accelerating Higher Education Expansion and Development (AHEAD) Operation of the Ministry of Higher Education of Sri Lanka funded by the World Bank (https://ahead.lk/result-area-3/). Publisher Copyright: © 2022 IEEE.

PY - 2022/12/20

Y1 - 2022/12/20

N2 - Magnitude-based pruning is a technique used to optimise deep learning models for edge inference. We have achieved over 75% model size reduction with a higher accuracy than the original multi-output regression model for head-pose estimation.

AB - Magnitude-based pruning is a technique used to optimise deep learning models for edge inference. We have achieved over 75% model size reduction with a higher accuracy than the original multi-output regression model for head-pose estimation.

KW - Edge inference

KW - Optimisation

KW - Network Pruning

KW - Quantisation

KW - TensorFlow

KW - Head Post estimation

KW - Head Pose estimation

KW - Edge Inference

UR - http://www.scopus.com/inward/record.url?scp=85145652121&partnerID=8YFLogxK

U2 - 10.1109/tencon55691.2022.9977637

DO - 10.1109/tencon55691.2022.9977637

M3 - Conference contribution

SN - 978-1-6654-5096-6

BT - Proceedings of TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON)

PB - IEEE

ER -

Comparative Study of Parameter Selection for Enhanced Edge Inference for a Multi-Output Regression model for Head Pose Estimation

Abstract

Publication series

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this