A Realtime Lightweight Detection Network Framework for UAV Identification

  • Guoqi Zeng
  • , Shengrui Pan
  • , Zheng Fan
  • , Guida Wang
  • , Ming Wang*
  • , Huanying Yue
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Detection Neural networks are memory intensive, making them difficult to deploy on embedded system with limited hardware resources. To solve the real-time problem of airborne recognition, we introduce a lightweight detection network framework based on the YOLOv3-tiny, that work together to reduce the volume of network by 3× with accuracy increasing by 1% and detection speed reaching 15 fps. The method is based on the YOLOv3-tiny, the size of which is 33.1 MB. Its reference time and mAP are 0.085 s and 78.6%, respectively. Next, we use the structure compression and improved channel pruning strategy to reduce the network volume to 10.33 MB without affecting their accuracy. Considering that model compression will have an impact on the performance of the network, we also adopt some tricks, such as adjusting learning rate and adding attention mechanism, to further improve the identification accuracy, from 78.6% to 79.3%. These measures enable the model to run better and faster on the resource-constrained airborne embedded devices or mobile applications.

Original languageEnglish
Title of host publicationProceedings of 2021 International Conference on Autonomous Unmanned Systems, ICAUS 2021
EditorsMeiping Wu, Yifeng Niu, Mancang Gu, Jin Cheng
PublisherSpringer Science and Business Media Deutschland GmbH
Pages2679-2688
Number of pages10
ISBN (Print)9789811694912
DOIs
StatePublished - 2022
EventInternational Conference on Autonomous Unmanned Systems, ICAUS 2021 - Changsha, China
Duration: 24 Sep 202126 Sep 2021

Publication series

NameLecture Notes in Electrical Engineering
Volume861 LNEE
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

ConferenceInternational Conference on Autonomous Unmanned Systems, ICAUS 2021
Country/TerritoryChina
CityChangsha
Period24/09/2126/09/21

Keywords

  • Channel domain attention module
  • Lightweight detection neural network
  • Model compression
  • UAV identification
  • YOLOv3-Tiny

Fingerprint

Dive into the research topics of 'A Realtime Lightweight Detection Network Framework for UAV Identification'. Together they form a unique fingerprint.

Cite this