首页>成果

  • 综合
  • 标题
  • 关键词
  • 摘要
  • 学者
  • 期刊-刊名
  • 期刊-ISSN
  • 会议名称
搜索

[会议论文]

Face detection under variable lighting based on resample by face relighting

分享
编辑 删除 报错

作者:

Li, YM (Li, YM.) (学者:李艳梅) | Chen, R (Chen, R.) | Qing, L (Qing, L.) | 展开

收录:

CPCI-S

摘要:

Different environment illuminations have a great impact on face detection. In this paper, we present a solution based on face relighting technology. The basic idea is that there exists nine harmonic images that can be derived from a 3D model of a face, and by which we can estimate the illumination coefficient of any face samples. Using an illumination radio image, we can produce images under new lighting conditions. To detect faces under certain lighting condition, we relight original face samples to get more new faces under kinds of possible lighting condition and add them to the training set. Our experimental results on Support Vector Machine (SVM) turn out that the relighting subspace is effective on detection under variations of the lighting conditions. Moreover, if we relight original face samples to new samples under different illuminations, the collected example sets will be multiplied. We use the expanded database to train an AdaBoost-based face detector and test it on the MIT+CMU frontal face test set. The experimental results show that the data collection can be efficiently speeded up by the proposed methods. The later experiment also verifies the generalization capability of the proposed method.

关键词:

refighting face detection harmonic images Lambertian surface illumination SVM

作者机构:

  • [ 1 ] Beijing Univ Technol, Multimedia & Intelligent Software Technol Lab, Beijing 100022, Peoples R China

通讯作者信息:

  • 李艳梅

    [Li, YM]Beijing Univ Technol, Multimedia & Intelligent Software Technol Lab, Beijing 100022, Peoples R China

查看成果更多字段

相关文章:

来源 :

PROCEEDINGS OF THE 2004 INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND CYBERNETICS, VOLS 1-7

年份: 2004

页码: 3775-3780

语种: 英文

被引次数:

WoS核心集被引频次: 4

近30日浏览量: 0

在线人数/总访问数:119/4548826
地址:北京工业大学图书馆(北京市朝阳区平乐园100号 邮编:100124) 联系我们:010-67392185
版权所有:北京工业大学图书馆 站点建设与维护:北京爱琴海乐之技术有限公司