Unsupervised Image-Generation Enhanced Adaptation for Object Detection in Thermal Images

doi:10.1155/2021/1837894

	Unsupervised Image-Generation Enhanced Adaptation for Object Detection in Thermal Images
	Liu, Peng 2; Li, Fuyu 1; Yuan, Shanshan 2; Li, Wanyi 1
刊名	MOBILE INFORMATION SYSTEMS
	2021-12-27
卷号	2021 页码:6
ISSN号	1574-017X
DOI	10.1155/2021/1837894
通讯作者	Li, Wanyi(wanyi.li@ia.ac.cn)
英文摘要	Object detection in thermal images is an important computer vision task and has many applications such as unmanned vehicles, robotics, surveillance, and night vision. Deep learning-based detectors have achieved major progress, which usually need large amount of labelled training data. However, labelled data for object detection in thermal images is scarce and expensive to collect. How to take advantage of the large number labelled visible images and adapt them into thermal image domain is expected to solve. This paper proposes an unsupervised image-generation enhanced adaptation method for object detection in thermal images. To reduce the gap between visible domain and thermal domain, the proposed method manages to generate simulated fake thermal images that are similar to the target images and preserves the annotation information of the visible source domain. The image generation includes a CycleGAN-based image-to-image translation and an intensity inversion transformation. Generated fake thermal images are used as renewed source domain, and then the off-the-shelf domain adaptive faster RCNN is utilized to reduce the gap between the generated intermediate domain and the thermal target domain. Experiments demonstrate the effectiveness and superiority of the proposed method.
资助项目	Science and Technology Plan Project of State Administration[2020 MK 162] ; National Natural Science Foundation of China[61771471] ; National Natural Science Foundation of China[61401463] ; National Natural Science Foundation of China[U1613213] ; National Natural Science Foundation of China[91748131] ; Central Foundational Research Funding Project[562020Y-7482]
WOS研究方向	Computer Science ; Telecommunications
语种	英语
出版者	HINDAWI LTD
WOS记录号	WOS:000802922500001
资助机构	Science and Technology Plan Project of State Administration ; National Natural Science Foundation of China ; Central Foundational Research Funding Project
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/49524]
专题	智能机器人系统研究
通讯作者	Li, Wanyi
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing, Peoples R China 2.China Natl Inst Standardizat, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Liu, Peng,Li, Fuyu,Yuan, Shanshan,et al. Unsupervised Image-Generation Enhanced Adaptation for Object Detection in Thermal Images[J]. MOBILE INFORMATION SYSTEMS,2021,2021:6.
APA	Liu, Peng,Li, Fuyu,Yuan, Shanshan,&Li, Wanyi.(2021).Unsupervised Image-Generation Enhanced Adaptation for Object Detection in Thermal Images.MOBILE INFORMATION SYSTEMS,2021,6.
MLA	Liu, Peng,et al."Unsupervised Image-Generation Enhanced Adaptation for Object Detection in Thermal Images".MOBILE INFORMATION SYSTEMS 2021(2021):6.