代替学習による次世代型知的防犯カメラのための視点変化に頑健なオブジェクト検知

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2019/09/01

タイトル(英語): A Study on Robust Object Recognition System for Intelligent Security Camera by using Alternative Deep Learning

著者名: 長山　格（琉球大学工学部知能情報コース），上原　和加貴（琉球大学大学院理工学研究科情報工学専攻），宮里　太也（琉球大学大学院理工学研究科情報工学専攻）

著者名(英語): Itaru Nagayama (Department of Information Engineering, University of the Ryukyus), Wakaki Uehara (Graduate School of Engineering, University of the Ryukyus), Takaya Miyazato (Graduate School of Engineering, University of the Ryukyus)

キーワード: 深層学習，ひったくり，防犯カメラ，代替学習，オブジェクト認識，3DCG　　Deep Learning，Snatching，Security Camera，Alternative Learning，Object Recognition，3DCG

要約(英語): An alternative learning and its application to construct a robust object recognition system for intelligent security camera(RORSIS) is presented in this paper. We show some experimental results of the development of the new robust 3-D object recognition system for intelligent security camera. In this system, a deep neural network and alternative learning using 3-D CG are key techniques for object recognition from a free viewpoint. Alternative learning is effective approach for the machine learning that depends on huge amount of tarining data. Some appearance based characteristics are determined from captured images, and the system uses a deep neural network, called AlexNet, to automatically classify bicycle, automobile and so on. The proposed system shows that several kinds of equipments can be recognized from a free view point. Experimental results show that the system can effectively recognize four kinds of real objects with 99.5% accuracy.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.139 No.9 （2019） 特集：知能メカトロニクス分野と連携する知覚情報技術

本誌掲載ページ: 964-971 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/139/9/139_964/_article/-char/ja/