Region Proposal by Guided Anchoring

被引:518
作者
Wang, Jiaqi [1 ]
Chen, Kai [1 ]
Yang, Shuo [2 ]
Loy, Chen Change [3 ]
Lin, Dahua [1 ]
机构
[1] Chinese Univ Hong Kong, CUHK SenseTime Joint Lab, Hong Kong, Peoples R China
[2] Chinese Univ Hong Kong, Hong Kong, Peoples R China
[3] Nanyang Technol Univ, Singapore, Singapore
来源
2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019) | 2019年
关键词
D O I
10.1109/CVPR.2019.00308
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Region anchors are the cornerstone of modern object detection techniques. State-of-the-art detectors mostly rely on a dense anchoring scheme, where anchors are sampled uniformly over the spatial domain with a predefined set of scales and aspect ratios. In this paper, we revisit this foundational stage. Our study shows that it can be done much more effectively and efficiently. Specifically, we present an alternative scheme, named Guided Anchoring, which leverages semantic features to guide the anchoring. The proposed method jointly predicts the locations where the center of objects of interest are likely to exist as well as the scales and aspect ratios at different locations. On top of predicted anchor shapes, we mitigate the feature inconsistency with a feature adaption module. We also study the use of high-quality proposals to improve detection performance. The anchoring scheme can be seamlessly integrated into proposal methods and detectors. With Guided Anchoring, we achieve 9.1% higher recall on MS COCO with 90% fewer anchors than the RPN baseline. We also adopt Guided Anchoring in Fast R-CNN, Faster R-CNN and RetinaNet, respectively improving the detection mAP by 2.2%, 2.7% and 1.2%. Code is available at https://github.com/open-mmlab/mmdetection.
引用
收藏
页码:2960 / 2969
页数:10
相关论文
共 32 条
[1]  
[Anonymous], 2017, ARXIV171010749
[2]  
[Anonymous], INT J COMPUTER VISIO
[3]  
Chen Kai, 2019, Hybrid task cascade for instance segmentation
[4]  
Dai J., 2016, ADV NEURAL INFORM PR, P379, DOI DOI 10.1109/CVPR.2017.690
[5]  
Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848
[6]   The PASCAL Visual Object Classes Challenge: A Retrospective [J].
Everingham, Mark ;
Eslami, S. M. Ali ;
Van Gool, Luc ;
Williams, Christopher K. I. ;
Winn, John ;
Zisserman, Andrew .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2015, 111 (01) :98-136
[7]  
Geiger A., 2012, C COMP VIS PATT REC
[8]   DeepProposal: Hunting Objects by Cascading Deep Convolutional Layers [J].
Ghodrati, Amir ;
Diba, Ali ;
Pedersoli, Marco ;
Tuytelaars, Tinne ;
Van Gool, Luc .
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2015, :2578-2586
[9]   Object detection via a multi-region & semantic segmentation-aware CNN model [J].
Gidaris, Spyros ;
Komodakis, Nikos .
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2015, :1134-1142
[10]  
Girshick R., 2015, P IEEE INT C COMPUTE, DOI [DOI 10.1109/ICCV.2015.169, 10.1109/ICCV.2015.169]