Weakly Supervised Object Detection via Object-Specific Pixel Gradient | |
Shen, Yunhang1; Ji, Rongrong1; Wang, Changhu2; Li, Xi3; Li, Xuelong4,5 | |
刊名 | IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS |
2018-12 | |
卷号 | 29期号:12页码:5960-5970 |
关键词 | Convolutional Neural Network (Cnn) Object Detection Weakly Supervised |
ISSN号 | 2162-237X;2162-2388 |
DOI | 10.1109/TNNLS.2018.2816021 |
产权排序 | 4 |
英文摘要 | Most existing object detection algorithms are trained based upon a set of fully annotated object regions or bounding boxes, which are typically labor-intensive. On the contrary, nowadays there is a significant amount of imagelevel annotations cheaply available on the Internet. It is hence a natural thought to explore such "weak" supervision to benefit the training of object detectors. In this paper, we propose a novel scheme to perform weakly supervised object localization, termed object-specific pixel gradient (OPG). The OPG is trained by using image-level annotations alone, which performs in an iterative manner to localize potential objects in a given image robustly and efficiently. In particular, we first extract an OPG map to reveal the contributions of individual pixels to a given object category, upon which an iterative mining scheme is further introduced to extract instances or components of this object. Moreover, a novel average and max pooling layer is introduced to improve the localization accuracy. In the task of weakly supervised object localization, the OPG achieves a state-of-the-art 44.5% top-5 error on ILSVRC 2013, which outperforms competing methods, including Oquab et al. and region-based convolutional neural networks on the Pascal VOC 2012, with gains of 2.6% and 2.3%, respectively. In the task of object detection, OPG achieves a comparable performance of 27.0% mean average precision on Pascal VOC 2007. In all experiments, the OPG only adopts the off-the-shelf pretrained CNN model, without using any object proposals. Therefore, it also significantly improves the detection speed, i.e., achieving three times faster compared with the stateof-the-art method. |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:000451230100014 |
内容类型 | 期刊论文 |
源URL | [http://ir.opt.ac.cn/handle/181661/30739] |
专题 | 西安光学精密机械研究所_光学影像学习与分析中心 |
通讯作者 | Ji, Rongrong |
作者单位 | 1.Xiamen Univ, Sch Informat Sci & Engn, Fujian Key Lab Sensing & Comp Smart City, Xiamen 361005, Peoples R China 2.Toutiao AI Lab, Beijing 100098, Peoples R China 3.Zhejiang Univ, Coll Comp Sci & Technol, Hangzhou 310027, Zhejiang, Peoples R China 4.Chinese Acad Sci, Xian Inst Opt & Precis Mech, Xian 710119, Shaanxi, Peoples R China 5.Univ Chinese Acad Sci, Beijing 100049, Peoples R China |
推荐引用方式 GB/T 7714 | Shen, Yunhang,Ji, Rongrong,Wang, Changhu,et al. Weakly Supervised Object Detection via Object-Specific Pixel Gradient[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,2018,29(12):5960-5970. |
APA | Shen, Yunhang,Ji, Rongrong,Wang, Changhu,Li, Xi,&Li, Xuelong.(2018).Weakly Supervised Object Detection via Object-Specific Pixel Gradient.IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,29(12),5960-5970. |
MLA | Shen, Yunhang,et al."Weakly Supervised Object Detection via Object-Specific Pixel Gradient".IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 29.12(2018):5960-5970. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论