Target Detection Label Assignment Method Based on Global Information

With the development of deep learning framework, new object detection algorithms have also been proposed, such as first-stage and two-stage detection models, which have improved the detection speed and solved the problem of object detection at different scales, but they have not yet been well solved...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHANG Pei-pei, LU Zhen-yu
Format: Article
Language:zho
Published: Harbin University of Science and Technology Publications 2022-08-01
Series:Journal of Harbin University of Science and Technology
Subjects:
Online Access:https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2114
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849303615828656128
author ZHANG Pei-pei
LU Zhen-yu
author_facet ZHANG Pei-pei
LU Zhen-yu
author_sort ZHANG Pei-pei
collection DOAJ
description With the development of deep learning framework, new object detection algorithms have also been proposed, such as first-stage and two-stage detection models, which have improved the detection speed and solved the problem of object detection at different scales, but they have not yet been well solved for overlapping, occlusion and other issues. One of the reasons for this problem is that during model training, label assignment is not done well. Aiming at this problem, this paper proposes a target detection label allocation method based on global information, which uses the assignment method to establish a global optimal label allocation mathematical model based on the loss function in the model training stage, and gives the fusion mode of the model with other object detection models, and the role played by the method in the process of object detection. The experimental results show that the detection model of the fusion method is better than that of the model that does not use the method under the complex detection scenarios such as overlapping and occlusion.
format Article
id doaj-art-c65362bfe199442f813d559fef4a0b7b
institution Kabale University
issn 1007-2683
language zho
publishDate 2022-08-01
publisher Harbin University of Science and Technology Publications
record_format Article
series Journal of Harbin University of Science and Technology
spelling doaj-art-c65362bfe199442f813d559fef4a0b7b2025-08-20T03:56:03ZzhoHarbin University of Science and Technology PublicationsJournal of Harbin University of Science and Technology1007-26832022-08-012704323810.15938/j.jhust.2022.04.005Target Detection Label Assignment Method Based on Global InformationZHANG Pei-pei0LU Zhen-yu1College of Management , North China University of Science and Technology, Tangshan 063000, ChinaCollege of Management , North China University of Science and Technology, Tangshan 063000, ChinaWith the development of deep learning framework, new object detection algorithms have also been proposed, such as first-stage and two-stage detection models, which have improved the detection speed and solved the problem of object detection at different scales, but they have not yet been well solved for overlapping, occlusion and other issues. One of the reasons for this problem is that during model training, label assignment is not done well. Aiming at this problem, this paper proposes a target detection label allocation method based on global information, which uses the assignment method to establish a global optimal label allocation mathematical model based on the loss function in the model training stage, and gives the fusion mode of the model with other object detection models, and the role played by the method in the process of object detection. The experimental results show that the detection model of the fusion method is better than that of the model that does not use the method under the complex detection scenarios such as overlapping and occlusion.https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2114object detectionlabel assignmentassign the modeldeep learning
spellingShingle ZHANG Pei-pei
LU Zhen-yu
Target Detection Label Assignment Method Based on Global Information
Journal of Harbin University of Science and Technology
object detection
label assignment
assign the model
deep learning
title Target Detection Label Assignment Method Based on Global Information
title_full Target Detection Label Assignment Method Based on Global Information
title_fullStr Target Detection Label Assignment Method Based on Global Information
title_full_unstemmed Target Detection Label Assignment Method Based on Global Information
title_short Target Detection Label Assignment Method Based on Global Information
title_sort target detection label assignment method based on global information
topic object detection
label assignment
assign the model
deep learning
url https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2114
work_keys_str_mv AT zhangpeipei targetdetectionlabelassignmentmethodbasedonglobalinformation
AT luzhenyu targetdetectionlabelassignmentmethodbasedonglobalinformation