Transductive Cartoon Retrieval by Multiple Hypergraph Learning

Publication Type:
Conference Proceeding
Lecture Notes in Computer Science, 2012, 7665 pp. 269 - 276
Issue Date:
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail2012004180OK.pdf739.17 kB
Adobe PDF
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.
Please use this identifier to cite or link to this item: