Interactive Regression and Classification for Dense Object Detector

doi:10.1109/TIP.2022.3174391

	Interactive Regression and Classification for Dense Object Detector
	Zhou, Linmao 1,2; Chang, Hong 1,2,3; Ma, Bingpeng 2; Shan, Shiguang 1
	2022
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
卷号	31 页码:3684-3696
摘要	In object detection, enhancing feature representation using localization information has been revealed as a crucial procedure to improve detection performance. However, the localization information (i.e., regression feature and regression offset) captured by the regression branch is still not well utilized. In this paper, we propose a simple but effective method called Interactive Regression and Classification (IRC) to better utilize localization information. Specifically, we propose Feature Aggregation Module (FAM) and Localization Attention Module (LAM) to leverage localization information to the classification branch during forward propagation. Furthermore, the classifier also guides the learning of the regression branch during backward propagation, to guarantee that the localization information is beneficial to both regression and classification. Thus, the regression and classification branches are learned in an interactive manner. Our method can be easily integrated into anchor-based and anchor-free object detectors without increasing computation cost. With our method, the performance is significantly improved on many popular dense object detectors, including RetinaNet, FCOS, ATSS, PAA, GFL, GFLV2, OTA, GA-RetinaNet, RepPoints, BorderDet and VFNet. Based on ResNet-101 backbone, IRC achieves 47.2% AP on COCO test-dev, surpassing the previous state-of-the-art PAA (44.8% AP), GFL (45.0% AP) and without sacrificing the efficiency both in training and inference. Moreover, our best model (Res2Net-101-DCN) can achieve a single-model single-scale AP of 51.4%.
关键词	Location awareness Detectors Feature extraction Object detection Standards Backpropagation Pipelines Dense object detector localization information interactive
DOI	10.1109/TIP.2022.3174391
收录类别	SCI
语种	英语
资助项目	Natural Science Foundation of China (NSFC)[61976203] ; Natural Science Foundation of China (NSFC)[61876171]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000803395500010
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：3[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/19587
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Chang, Hong
作者单位	1.Chinese Acad Sci, Key Lab Intelligent Informat Proc, Inst Comp Technol CAS, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Comp Sci & Technol, Beijing 100049, Peoples R China 3.Peng Cheng Lab, Shenzhen 518055, Peoples R China
推荐引用方式 GB/T 7714	Zhou, Linmao,Chang, Hong,Ma, Bingpeng,et al. Interactive Regression and Classification for Dense Object Detector[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2022,31:3684-3696.
APA	Zhou, Linmao,Chang, Hong,Ma, Bingpeng,&Shan, Shiguang.(2022).Interactive Regression and Classification for Dense Object Detector.IEEE TRANSACTIONS ON IMAGE PROCESSING,31,3684-3696.
MLA	Zhou, Linmao,et al."Interactive Regression and Classification for Dense Object Detector".IEEE TRANSACTIONS ON IMAGE PROCESSING 31(2022):3684-3696.