TransWeaver: Weave Image Pairs for Class Agnostic Common Object Detection

doi:10.1109/TIP.2023.3275870

	TransWeaver: Weave Image Pairs for Class Agnostic Common Object Detection
	Guo, Xiaoqian 1,2; Li, Xiangyang 1,2; Wang, Yaowei 3; Jiang, Shuqiang 1,2,3
	2023
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
卷号	32 页码:2947-2959
摘要	Measuring the similarity of two images is of crucial importance in computer vision. Class agnostic common object detection is a nascent research topic about mining image similarity, which aims to detect common object pairs from two images without category information. This task is general and less restrictive which explores the similarity between objects and can further describe the commonality of image pairs at the object level. However, previous works suffer from features with low discrimination caused by the lack of category information. Moreover, most existing methods compare objects extracted from two images in a simple and direct way, ignoring the internal relationships between objects in the two images. To overcome these limitations, in this paper, we propose a new framework called TransWeaver, which learns intrinsic relationships between objects. Our TransWeaver takes image pairs as input and flexibly captures the inherent correlation between candidate objects from two images. It consists of two modules (i.e., the representation-encoder and the weave-decoder) and captures efficient context information by weaving image pairs to make them interact with each other. The representation-encoder is used for representation learning, which can obtain more discriminative representations for candidate proposals. Furthermore, the weave-decoder weaves the objects from two images and is able to explore the inter-image and intra-image context information at the same time, bringing a better object matching ability. We reorganize the PASCAL VOC, COCO, and Visual Genome datasets to obtain training and testing image pairs. Extensive experiments demonstrate the effectiveness of the proposed TransWeaver which achieves state-of-the-art performance on all datasets.
关键词	Proposals Object detection Task analysis Feature extraction Visualization Training Measurement Common object detection transweaver transformer
DOI	10.1109/TIP.2023.3275870
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62125207] ; National Natural Science Foundation of China[62102400] ; National Natural Science Foundation of China[U19B2040] ; National Natural Science Foundation of China[U20B2052] ; National Postdoctoral Program for Innovative Talents[BX20200338] ; Beijing Natural Science Foundation[Z190020]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000995885700006
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/21208
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Jiang, Shuqiang
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 3.Peng Cheng Lab, Shenzhen 518055, Peoples R China
推荐引用方式 GB/T 7714	Guo, Xiaoqian,Li, Xiangyang,Wang, Yaowei,et al. TransWeaver: Weave Image Pairs for Class Agnostic Common Object Detection[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2023,32:2947-2959.
APA	Guo, Xiaoqian,Li, Xiangyang,Wang, Yaowei,&Jiang, Shuqiang.(2023).TransWeaver: Weave Image Pairs for Class Agnostic Common Object Detection.IEEE TRANSACTIONS ON IMAGE PROCESSING,32,2947-2959.
MLA	Guo, Xiaoqian,et al."TransWeaver: Weave Image Pairs for Class Agnostic Common Object Detection".IEEE TRANSACTIONS ON IMAGE PROCESSING 32(2023):2947-2959.