Abstract

Aerial object detection on an UAV or embedded vision platform requires accurate detection of objects with various spatial scales and has numerous applications in surveillance, traffic monitoring, search, and rescue, etc. The task of small-object detection becomes harder while using standard convolutional neural network architectures due to the reduction in spatial resolution. This work evaluates the effectiveness of using feature pyramid hierarchies with the Faster R-CNN algorithm for aerial object detection. The VisDrone aerial object detection dataset with ten object classes has been utilized to develop a Faster R-CNN ResNet model with C4 and FPN architectures to compare the performance. Significant improvement in the performance obtained by using feature pyramid networks for all object categories highlights their importance in the multi-scale aerial object detection task.

Original languageEnglish
Title of host publicationInformation and Communication Technology for Competitive Strategies, ICTCS 2021- ICT
Subtitle of host publicationApplications and Social Interfaces
EditorsAmit Joshi, Mufti Mahmud, Roshan G. Ragel
PublisherSpringer Science and Business Media Deutschland GmbH
Pages303-313
Number of pages11
ISBN (Print)9789811900945
DOIs
Publication statusPublished - 2023
Event6th International Conference on Information and Communication Technology for Competitive Strategies, ICTCS 2021 - Jaipur, India
Duration: 17-12-202118-12-2021

Publication series

NameLecture Notes in Networks and Systems
Volume400
ISSN (Print)2367-3370
ISSN (Electronic)2367-3389

Conference

Conference6th International Conference on Information and Communication Technology for Competitive Strategies, ICTCS 2021
Country/TerritoryIndia
CityJaipur
Period17-12-2118-12-21

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Signal Processing
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Multi-scale Aerial Object Detection Using Feature Pyramid Networks'. Together they form a unique fingerprint.

Cite this