文献类型：期刊文献

英文题名：Improved Target Detection Algorithm Based on Libra R-CNN

作者：Zhao, Zijing[1];Li, Xuewei[1];Liu, Hongzhe[1];Xu, Cheng[1]

第一作者：Zhao, Zijing

通讯作者：Li, XW[1]

机构：[1]Beijing Union Univ, Beijing Key Lab Informat Serv Engn, Beijing 100101, Peoples R China

第一机构：北京联合大学北京市信息服务工程重点实验室

通讯机构：[1]corresponding author), Beijing Union Univ, Beijing Key Lab Informat Serv Engn, Beijing 100101, Peoples R China.|[11417103]北京联合大学北京市信息服务工程重点实验室;[11417]北京联合大学;

年份：2020

卷号：8

起止页码：114044-114056

外文期刊名：IEEE ACCESS

收录：;EI(收录号：20202808925837);Scopus(收录号：2-s2.0-85087610803);WOS:【SCI-EXPANDED(收录号:WOS:000546417900026)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 61871039, Grant 61802019, and Grant 61906017, in part by the Supporting Plan for Cultivating High Level Teachers in Colleges and Universities in Beijing under Grant IDHT20170511, in part by the Premium Funding Project for Academic Human Resources Development with Beijing Union University under Grant BPHR2019AZ01, in part by the Beijing Municipal Commission of Education Project under Grant KM201911417001, in part by the Big Data Collaborative Innovation Centre for Intelligent Driving under Grant CYXC1902, in part by the Beijing Union University Project under Grant ZK80202001, Grant 202011417004, Grant 202011417005, and Grant 202011417SJ025, in part by the Project of Science and Technology Research and Development Plan of China National Railway Group Company Ltd., under Grant K2019Z006, and in part by the Graduate Program of Beijing Union University.

语种：英文

外文关键词：Object detection; Training; Feature extraction; Classification algorithms; Shape; Deep learning; Deep learning; target detection; traffic sign detection; improved Libra R-CNN

摘要：With the development of science and technology, artificial intelligence has been widely used in the transportation field, and research on the symmetry of artificial intelligence has become increasingly more in-depth. Traffic sign detection based on deep learning has the problems of different target shapes and high variability in the number of targets between different labels. To solve these problems from a lack of symmetry, the idea of applying the concept of balanced data and the deformable positioning region to a target recognition network is proposed. The research is based on the improvement of the Libra R-CNN. Aiming at the problem that the difficult-to-distinguish target in target detection has a high impact on detection, the idea of generating increasingly more diverse indistinguishable samples during training is proposed to improve the detection accuracy, which is verified by experiments. The experiment is carried out on the MS COCO 2017 and traffic sign datasets. The improved Libra R-CNN is 3 percentage points better than the unimproved Libra R-CNN's mean Average Precision (mAP). A large number of comparative experimental results show that the improved network is effective.