Weakly Supervised Large Scale Object Localization with Multiple Instance Learning and Bag Splitting
文献类型：期刊论文
| 作者 | Ren, Weiqiang1,2; Huang, Kaiqi1,2 ; Tao, Dacheng3,4; Tan, Tieniu1,2
|
| 刊名 | IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
 |
| 出版日期 | 2016-02-01 |
| 卷号 | 38期号:2页码:405-416 |
| 关键词 | Weakly Supervised Localization Convolutional Networks Multiple Instance Learning |
| DOI | 10.1109/TPAMI.2015.2456908 |
| 文献子类 | Article |
| 英文摘要 | Localizing objects of interest in images when provided with only image-level labels is a challenging visual recognition task. Previous efforts have required carefully designed features and have difficulty in handling images with cluttered backgrounds. Up-scaling to large datasets also poses a challenge to applying these methods to real applications. In this paper, we propose an efficient and effective learning framework called MILinear, which is able to learn an object localization model from large-scale data without using bounding box annotations. We integrate rich general prior knowledge into a learning model using a large pre-trained convolutional network. Moreover, to reduce ambiguity in positive images, we present a bag-splitting algorithm that iteratively generates new negative bags from positive ones. We evaluate the proposed approach on the challenging Pascal VOC 2007 dataset, and our method outperforms other state-of-the-art methods by a large margin; some results are even comparable to fully supervised models trained with bounding box annotations. To further demonstrate scalability, we also present detection results on the ILSVRC 2013 detection dataset, and our method outperforms supervised deformable part-based model without using box annotations. |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:000369989600016 |
| 资助机构 | National Basic Research Program of China(2012CB316302) ; National Natural Science Foundation of China(61322209 ; China Scholarship Council ; Australian Research Council(DP-140102164 ; 61175007) ; FT-130101457) |
| 源URL | [http://ir.ia.ac.cn/handle/173211/11342]  |
| 专题 | 自动化研究所_智能感知与计算研究中心 |
| 作者单位 | 1.Chinese Acad Sci CASIA, Ctr Res Intelligent Percept & Comp CRIPAC, 95 ZhongGuanCun East St, Beijing 100190, Peoples R China 2.Chinese Acad Sci CASIA, Inst Automat, Natl Lab Pattern Recognit NLPR, 95 ZhongGuanCun East St, Beijing 100190, Peoples R China 3.Univ Technol Sydney, Ctr Quantum Computat & Intelligent Syst, 81 Broadway St, Ultimo, NSW 2007, Australia 4.Univ Technol Sydney, Fac Engn & Informat Technol, 81 Broadway St, Ultimo, NSW 2007, Australia |
| 推荐引用方式 GB/T 7714 | Ren, Weiqiang,Huang, Kaiqi,Tao, Dacheng,et al. Weakly Supervised Large Scale Object Localization with Multiple Instance Learning and Bag Splitting[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,2016,38(2):405-416. |
| APA | Ren, Weiqiang,Huang, Kaiqi,Tao, Dacheng,&Tan, Tieniu.(2016).Weakly Supervised Large Scale Object Localization with Multiple Instance Learning and Bag Splitting.IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,38(2),405-416. |
| MLA | Ren, Weiqiang,et al."Weakly Supervised Large Scale Object Localization with Multiple Instance Learning and Bag Splitting".IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 38.2(2016):405-416. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
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。