Image Clustering Based on Multi-Scale Deep Maximize Mutual Information and Self-Training Algorithm | |
Yu SQ(余思泉)2,3,4; Shen GP(沈贵萍)5; Wang PY(王佩瑶)1; Wang YN(王宇宁)6; Wu CD(吴成东)7 | |
刊名 | IEEE ACCESS |
2020 | |
卷号 | 8页码:160285-160296 |
关键词 | Clustering algorithms Mutual information Image representation Neural networks Clustering methods Feature extraction Task analysis Representation learning image clustering mutual information maximization self-training algorithm |
ISSN号 | 2169-3536 |
产权排序 | 1 |
英文摘要 | Image clustering is a complex procedure that is significantly affected by the choice of the image representation. Generally speaking, image representations are generated by using handcraft features or trained neural networks. When dealing with high dimension data, these two representation methods cause two problems: i) the representation ability of the manually designed features is limited; ii) the non-representative and meaningless feature of a trained deep network may hurt the clustering performance. To overcome these problems, we propose a new clustering method which efficiently builds an image representation and precisely discovers the cluster assignments. Our main tools are an unsupervised representation learning method based on Deep Mutual Information Maximization (DMIM) system, and a clustering method based on self-training algorithm. Specifically speaking, to extract the informative representation of image data, we derive the maximum mutual information theory and propose a system to learn the maximum mutual information between the input images and the latent representations. To discover the clusters and assign each image a clustering label, a self-training mechanism is applied to cluster the learned representations. The superiority and validity of our algorithm are verified in a series of real-world image clustering experiments. |
WOS研究方向 | Computer Science ; Engineering ; Telecommunications |
语种 | 英语 |
WOS记录号 | WOS:000571046600001 |
内容类型 | 期刊论文 |
源URL | [http://ir.sia.cn/handle/173321/27678] |
专题 | 沈阳自动化研究所_机器人学研究室 |
通讯作者 | Shen GP(沈贵萍) |
作者单位 | 1.Shenyang Institute of Technology, Shenyang 110000, China 2.College of Information Science and Engineering, Northeastern University, Shenyang 110819, China 3.State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110169, China 4.Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang 110169, China 5.College of Arti cial Intelligence, Yango University, Fuzhou 350108, China 6.PLA 32681, Shenyang 110000, China 7.Faculty of Robot Science and Engineering, Northeastern University, Shenyang 110819, China |
推荐引用方式 GB/T 7714 | Yu SQ,Shen GP,Wang PY,et al. Image Clustering Based on Multi-Scale Deep Maximize Mutual Information and Self-Training Algorithm[J]. IEEE ACCESS,2020,8:160285-160296. |
APA | Yu SQ,Shen GP,Wang PY,Wang YN,&Wu CD.(2020).Image Clustering Based on Multi-Scale Deep Maximize Mutual Information and Self-Training Algorithm.IEEE ACCESS,8,160285-160296. |
MLA | Yu SQ,et al."Image Clustering Based on Multi-Scale Deep Maximize Mutual Information and Self-Training Algorithm".IEEE ACCESS 8(2020):160285-160296. |
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