著者

長山 格 上原 和加貴 城間 康 宮里 太也

出版者

一般社団法人 電気学会

雑誌

電気学会論文誌D(産業応用部門誌) (ISSN:09136339)

巻号頁・発行日

vol.141, no.2, pp.130-137, 2021

---

<p>In this study, we aim to develop a robust motion recognition system for an intelligent video surveillance system, that can be used for security, sports and rehabilitation by using extended alternative learning. A robust motion recognition system is necessary for the automated detection of security incidents by using a machine learning approach. However, to avoid the difficulty of collecting a huge training dataset, we propose an alternative learning approach that trains a deep neural network with a 3D-CG dataset to recognize several motions. We present our experimental results on motion recognition from free-viewpoint videos by using deep learning and alternative learning. The trained deep neural network (DNN) is evaluated using actual videos by classifying the different actions performed by real humans in these videos.</p>

著者

長山 格 岩永 竜也 上原 和加貴 宮里 太也

出版者

一般社団法人 電気学会

雑誌

電気学会論文誌D(産業応用部門誌) (ISSN:09136339)

巻号頁・発行日

vol.141, no.2, pp.138-146, 2021

---

<p>This paper presents the development of a new method for the estimation and resolution of body occlusion using deep learning for an advanced intelligent video surveillance system. A generative adversarial network is used to estimate and reconstruct an image of a hidden part of the human body. Furthermore, an alternative learning approach using 3DCG that was developed in our previous study is adopted to create a large dataset for deep learning. Experimental results indicate that the proposed method performs well in the estimation of hidden parts of the human body using images of actual people.</p>

著者

宮里 太也 上原 和加貴 長山 格

出版者

一般社団法人 電気学会

雑誌

電気学会論文誌C(電子・情報・システム部門誌) (ISSN:03854221)

巻号頁・発行日

vol.139, no.9, pp.972-979, 2019

---

<p>This paper describes the development of 3-D human recognition system of flying drone system for emergency rescue and investigation. In this system, deep neural network and its application for 3-D object recognition are key techniques for human detection from a free viewpoint. Some appearance based characteristics are captured as movie frames, and the system uses deep neural networks to automatically classify concerned object. The proposed system performs well that many kinds of views of personnels can be recognized from bird's-eye view. Experimental results show that the system can effectively recognize human objects with high accuracy.</p>

著者

長山 格 上原 和加貴 宮里 太也

出版者

一般社団法人 電気学会

雑誌

電気学会論文誌C(電子・情報・システム部門誌) (ISSN:03854221)

巻号頁・発行日

vol.139, no.9, pp.964-971, 2019

被引用文献数

1

---

<p>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.</p>