Image-based kinship recognition is an important research area in the social network. Beyond kinship verification, family relationship recognition is a challenging task with little existing literature. In this paper, we propose a pictorial structure based method for automatic family tree construction. Unlike traditional work based on a single image, an album-oriented method can leverage the cooperation of multiple instances associated with the same identity. In this way, the reliability of kinship similarity can be improved. Besides, the family tree structure is used to describe kinship relations among each pair of members. The experimental result shows that the proposed method can construct a family tree effectively and improve the overall accuracy of kinship identification.
CITATION STYLE
Xia, C., Xia, S., Shao, M., & Fu, Y. (2019). Album to Family Tree: A Graph Based Method for Family Relationship Recognition. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11362 LNCS, pp. 496–510). Springer Verlag. https://doi.org/10.1007/978-3-030-20890-5_32
Mendeley helps you to discover research relevant for your work.