Modality-specific and hierarchical feature learning for RGB-D hand-held object recognition

doi:10.1007/s11042-016-3375-5

CORC > 计算技术研究所 > 中国科学院计算技术研究所 > 中国科学院计算技术研究所期刊论文 > 英文

	Modality-specific and hierarchical feature learning for RGB-D hand-held object recognition
	Lv, Xiong 2; Liu, Xinda 3; Li, Xiangyang 2; Li, Xue 2,4; Jiang, Shuqiang 2; He, Zhiqiang 1
刊名	MULTIMEDIA TOOLS AND APPLICATIONS
	2017-02-01
卷号	76 期号:3 页码:4273-4290
关键词	Feature learning RGB-D object recogntion Multiple modalities
ISSN号	1380-7501
DOI	10.1007/s11042-016-3375-5
英文摘要	Hand-held object recognition is an important research topic in image understanding and plays an essential role in human-machine interaction. With the easily available RGB-D devices, the depth information greatly promotes the performance of object segmentation and provides additional channel information. While how to extract a representative and discriminating feature from object region and efficiently take advantage of the depth information plays an important role in improving hand-held object recognition accuracy and eventual human-machine interaction experience. In this paper, we focus on a special but important area called RGB-D hand-held object recognition and propose a hierarchical feature learning framework for this task. First, our framework learns modality-specific features from RGB and depth images using CNN architectures with different network depth and learning strategies. Secondly a high-level feature learning network is implemented for a comprehensive feature representation. Different with previous works on feature learning and representation, the hierarchical learning method can sufficiently dig out the characteristics of different modal information and efficiently fuse them in a unified framework. The experimental results on HOD dataset illustrate the effectiveness of our proposed method.
资助项目	National Basic Research 973 Program of China[2012CB316400] ; National Natural Science Foundation of China[61532018] ; National Natural Science Foundation of China[61322212] ; National High Technology Research and Development 863 Program of China[2014AA015202] ; Lenovo Outstanding Young Scientists Program (LOYS)
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	SPRINGER
WOS记录号	WOS:000396051200054
内容类型	期刊论文
源URL	[http://119.78.100.204/handle/2XEOYT63/7478]
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	He, Zhiqiang
作者单位	1.Lenovo Corp Res, Beijing 100085, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 3.Ningxia Univ, Sch Math & Comp Sci, Ningxia 750021, Peoples R China 4.Shandong Univ Sci & Technol, Coll Informat Sci & Engn, Qingdao, Shandong, Peoples R China
推荐引用方式 GB/T 7714	Lv, Xiong,Liu, Xinda,Li, Xiangyang,et al. Modality-specific and hierarchical feature learning for RGB-D hand-held object recognition[J]. MULTIMEDIA TOOLS AND APPLICATIONS,2017,76(3):4273-4290.
APA	Lv, Xiong,Liu, Xinda,Li, Xiangyang,Li, Xue,Jiang, Shuqiang,&He, Zhiqiang.(2017).Modality-specific and hierarchical feature learning for RGB-D hand-held object recognition.MULTIMEDIA TOOLS AND APPLICATIONS,76(3),4273-4290.
MLA	Lv, Xiong,et al."Modality-specific and hierarchical feature learning for RGB-D hand-held object recognition".MULTIMEDIA TOOLS AND APPLICATIONS 76.3(2017):4273-4290.