• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Artificial Intelligence: Technologies and Applications >A Lifelong Learning Approach for Improving Accurate Face Recognition
【24h】

A Lifelong Learning Approach for Improving Accurate Face Recognition

机译：一种提高精确面部识别的终身学习方法

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

With the popularity of artificial intelligence and computer vision, an increasing number of software engineers attempt to make their systems be able to recognize the user, through the way of face recognition, e.g., Characteristic Points (i.e., CP)-based face recognition. Generally, the traditional CPbased face recognition only work when the user's face is not much different from the one that stores in the system. However, users' faces can change a lot intentionally or unintentionally, which brings a great challenge for correct face recognition. In view of this challenge, a novel face recognition approach LL-FR (Lifelong Learning-based Face Recognition) is put forward in this paper. Concretely, the system stores not only the face image that the user registered, but also the images every time when it recognizes the user's face. Afterwards, the system makes the future recognition of the user based on all the previous faces of him/her. Finally, through a set of experiments, we demonstrate the feasibility of our approach.
机译：随着人工智能和计算机视觉的普及，越来越多的软件工程师试图使他们的系统能够识别用户，通过面部识别，例如，特征点（即CP）基于面部识别的方式。一般情况下，当用户的脸部是不是从一个非常不同的传统CPbased人脸识别只能工作在系统存储。但是，用户的面部可以改变很多有意还是无意，这带来了正确的面部识别一个巨大的挑战。鉴于这一挑战，一种新的人脸识别方法LL-FR（基于终身学习的人脸识别）在本文中被提出来的。具体而言，该系统不仅存储的面部图像，用户注册，而且在每次开机时识别用户的面部时的图像。然后，系统基于他以前所有的面孔/她的用户的未来肯定。最后，通过一组实验中，我们证明了该方法的可行性。

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