Correlational examples for convolutional neural networks to detect small impurities

doi:10.1016/j.neucom.2018.03.017

	Correlational examples for convolutional neural networks to detect small impurities
	Guo, Yue1,2 ; He, Yijia1,2 ; Song, Haitao1 ; He, Wenhao1 ; Yuan, Kui1
发表期刊	NEUROCOMPUTING
	2018-06-21
卷号	295 期号:21 页码:127-141
文章类型	Article
摘要	Convolutional neural networks have been significantly improving common object detection performances for a long time. However, targets across frames are independently detected in an image sequence, and object detection methods in multiple frames are generally divided into two main stages: object detection in every single frame and feature map association across frames. In this paper, a multi-frame detection framework is proposed to directly detect small impurities in opaque glass bottles with liquor. Specifically, a convolutional neural network trained with correlational examples simultaneously detects and correlates proposals, and then links them selectively to obtain robust detection results under challenging illuminations. Besides, memory costs of patch pairs become extremely large compared with those of patches, thus a sequential training procedure is introduced to relax hardware requirements. Extensive experiments on impurity datasets demonstrate superior performances of multi-frame detection frameworks with convolutional neural networks than traditional single-frame models. (C) 2018 Elsevier B.V. All rights reserved.
关键词	Impurity Detection Multi-frame Correlation Convolutional Neural Network Correlational Example Sequential Training
WOS标题词	Science & Technology ; Technology
DOI	10.1016/j.neucom.2018.03.017
收录类别	SCI
语种	英语
项目资助者	National Natural Science Foundation (NNSF) of China(61421004)
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:000430227300012
引用统计	被引频次：7[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/20969
专题	智能制造技术与系统研究中心_智能机器人
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing, Peoples R China 2.Univ Chinese Acad Sci, Sch Comp & Control Engn, Beijing, Peoples R China
第一作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Guo, Yue,He, Yijia,Song, Haitao,et al. Correlational examples for convolutional neural networks to detect small impurities[J]. NEUROCOMPUTING,2018,295(21):127-141.
APA	Guo, Yue,He, Yijia,Song, Haitao,He, Wenhao,&Yuan, Kui.(2018).Correlational examples for convolutional neural networks to detect small impurities.NEUROCOMPUTING,295(21),127-141.
MLA	Guo, Yue,et al."Correlational examples for convolutional neural networks to detect small impurities".NEUROCOMPUTING 295.21(2018):127-141.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
neurocomputing 2018.（6876KB）	期刊论文	作者接受稿	开放获取	CC BY-NC-SA	浏览