Melola Discovery

← Melola Discovery

Open access Journal SINTA 1

Deep learning and quantization for accurate and efficient multi-target radar inference of moving targets

International Journal of Electrical and Computer Engineering (IJECE) · Vol. 14 No. 3 (2024)

DOI
10.11591/ijece.v14i3.pp3187-3196
ISSN
2722-2578, 2088-8708
Published
2024-06-01
Publisher
Institute of Advanced Engineering and Science
License
creativecommons.org/licenses/by-sa/4.0

Abstract

Real-time, radar-based human activity and target classification is useful for wide-area ground surveillance. However, the feasibility of deploying deep learning (DL) models in radar-based systems with limited computational resources remains unexplored. This paper investigated the effect of quantization on model throughput and accuracy for deployment in radar systems. A seven-layer residual network was proposed to classify ground-moving targets and achieved a test accuracy of 87.72%. The model was then quantized to 16-bit and 8-bit precision, resulting in a 3.8 times speedup in inference throughput, with less than a 0.4% drop in test and validation accuracy. The results showed that quantization can improve inference throughput with a negligible decrease in target classification accuracy. The increase in throughput and reduction in computational expense that comes with quantization promotes the feasibility of the deployment of DL models in systems with limited computational resources. The findings of this paper hold significant promise for the successful use of quantized models in modern radar systems, while adhering to stringent size, weight and power consumption constraints.

Collaboration network of the authors

4 authors · 6 co-authorship links · 3 across different organisations

Hover or tap a name to highlight its co-authors. Double-click to open the author's page.

  • University of Cape Town
  • Council for Scientific and Industrial Research
  • Different organisations
  • Same organisation
  • Affiliation unknown
  • Thicker line = more joint articles · bigger circle = more articles
Show as table
Co-authorOrganisationJoint articles

Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.

This record comes from Crossref (what the publisher deposited), because the journal's own OAI-PMH endpoint could not be harvested.

Metadata source: Crossref API (cadangan, berdasarkan ISSN) https://api.crossref.org/journals/2088-8708/works, first harvested 27 Oct 2026, last seen 15 Nov 2026.
Metadata as oai_dc, converted from Crossref
<oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/"><dc:title>Deep learning and quantization for accurate and efficient multi-target radar inference of moving targets</dc:title><dc:creator>Ernest Mashanda, Nyasha</dc:creator><dc:creator>Watson, Neil</dc:creator><dc:creator>Berndt, Robert</dc:creator><dc:creator>Abdul Gaffar, Mohammed Yunus</dc:creator><dc:description>&lt;jats:p&gt;Real-time, radar-based human activity and target classification is useful for wide-area ground surveillance. However, the feasibility of deploying deep learning (DL) models in radar-based systems with limited computational resources remains unexplored. This paper investigated the effect of quantization on model throughput and accuracy for deployment in radar systems. A seven-layer residual network was proposed to classify ground-moving targets and achieved a test accuracy of 87.72%. The model was then quantized to 16-bit and 8-bit precision, resulting in a 3.8 times speedup in inference throughput, with less than a 0.4% drop in test and validation accuracy. The results showed that quantization can improve inference throughput with a negligible decrease in target classification accuracy. The increase in throughput and reduction in computational expense that comes with quantization promotes the feasibility of the deployment of DL models in systems with limited computational resources. The findings of this paper hold significant promise for the successful use of quantized models in modern radar systems, while adhering to stringent size, weight and power consumption constraints.&lt;/jats:p&gt;</dc:description><dc:publisher>Institute of Advanced Engineering and Science</dc:publisher><dc:date>2024-06-01</dc:date><dc:type>info:eu-repo/semantics/article</dc:type><dc:format>application/pdf</dc:format><dc:identifier>https://doi.org/10.11591/ijece.v14i3.pp3187-3196</dc:identifier><dc:identifier>10.11591/ijece.v14i3.pp3187-3196</dc:identifier><dc:source>International Journal of Electrical and Computer Engineering (IJECE); Vol 14 No 3 (2024); 3187</dc:source><dc:source>2722-2578</dc:source><dc:source>2088-8708</dc:source><dc:relation>https://ijece.iaescore.com/index.php/IJECE/article/viewFile/33385/17416</dc:relation><dc:rights>http://creativecommons.org/licenses/by-sa/4.0</dc:rights></oai_dc:dc>