Transductive cartoon retrieval by multiple hypergraph learning | |
Yu, Jun ; Cheng, J ; Wang, Jianmin ; Tao, Dacheng ; Yu J(俞俊) | |
2012 | |
关键词 | Data processing Estimation |
英文摘要 | Conference Name:19th International Conference on Neural Information Processing, ICONIP 2012. Conference Address: Doha, Qatar. Time:November 12, 2012 - November 15, 2012.; United Development Company PSC (UDC); Qatar Petrochemical Company; ExxonMobil; Qatar Petroleum; Texas A and M University at Qatar; Asia Pacific Neural Network Assembly; Cartoon characters retrieval frequently suffers from the distance estimation problem. In this paper, a multiple hypergraph fusion based approach is presented to solve this problem. We build multiple hypergraphs on cartoon characters based on their features. In these hypergraphs, each vertex is a character, and an edge links to multiple vertices. In this way, the distance estimation between characters is avoided and the high-order relationship among characters can be explored. The experiments of retrieval are conducted on cartoon datasets, and the results demonstrate that the proposed approach can achieve better performance than state-of-the-arts methods. 漏 2012 Springer-Verlag. |
语种 | 英语 |
出处 | http://dx.doi.org/10.1007/978-3-642-34487-9_33 |
出版者 | Springer Verlag |
内容类型 | 其他 |
源URL | [http://dspace.xmu.edu.cn/handle/2288/86765] |
专题 | 信息技术-会议论文 |
推荐引用方式 GB/T 7714 | Yu, Jun,Cheng, J,Wang, Jianmin,et al. Transductive cartoon retrieval by multiple hypergraph learning. 2012-01-01. |
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