講演抄録/キーワード |
講演名 |
2020-12-11 13:40
Study of Deep Learning based Object Detection for Automatic Train Operation in Railways ○Shiva Krishna Maheshuni(UTokyo)・Shimura Takahiro・Yabuki Kohei・Hasegawa Takumi(Kyosan Electric Mfg)・Takeshi Mizuma(UTokyo) DC2020-61 |
抄録 |
(和) |
(まだ登録されていません) |
(英) |
Real time object detection is already being implemented in systems like autonomous driving in cars, surveillance Cameras, factories, etc. But a very little application is being done in the field of railways. There are already various State of the Art object detectors which are trained on general open source image datasets like Open-Images 2019, MS-COCO , etc. The same object detectors cannot be used in the view point of railways. The aim of this paper is to refine and develop a YOLOv4 Object detector , which can aid in automatic train operation. |
キーワード |
(和) |
/ / / / / / / |
(英) |
Object Detection / Convolutional neural networks(CNN) / Deep learning / YOLOv4 / Automatic Train Operation / / / |
文献情報 |
信学技報, vol. 120, no. 288, DC2020-61, pp. 12-17, 2020年12月. |
資料番号 |
DC2020-61 |
発行日 |
2020-12-04 (DC) |
ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
DC2020-61 |
|