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 proceedingConference contributionpeer-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 languageEnglish
Title of host publicationProceedings of TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON)
PublisherIEEE
ISBN (Electronic)978-1-6654-5095-9
ISBN (Print)978-1-6654-5096-6
DOIs
Publication statusPublished (in print/issue) - 20 Dec 2022

Publication series

Name
PublisherIEEE 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

Cite this