首页> 外文会议>Asian Conference on Computer Vision >Album to Family Tree: A Graph Based Method for Family Relationship Recognition
【24h】

Album to Family Tree: A Graph Based Method for Family Relationship Recognition

机译:专辑到家族树:基于图的家族关系识别方法

获取原文

摘要

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.
机译:基于图像的亲属识别是社交网络中重要的研究领域。除了亲缘关系验证以外,很少有文献记载,家庭关系识别也是一项具有挑战性的任务。在本文中,我们提出了一种基于图画结构的自动家族树构建方法。与基于单个图像的传统作品不同,面向专辑的方法可以利用与同一身份关联的多个实例的协作。这样,可以提高亲属相似度的可靠性。此外,族谱结构用于描述每对成员之间的亲属关系。实验结果表明,该方法可以有效地构建家谱,提高亲属识别的整体准确性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号