Face Recognition Using an Adaptive Distance Isomap Approach

摘要：

Isomap as an efficient algorithm of manifold learning achieved in dimensionality reduction. It may not consider though the classification viewpoint. An adaptive distance Isomap approach for face recognition is presented. By Combining Isomap with an adaptive distance factor for classification, the distance between each two neighbor points will adaptively decreases, and points in the same class will be gathered. The experiments show that the recognition accuracy of the presented approach is greatly improved.

关键词： Isomap manifold adaptive distance Face recognition

作者: Lianzhao Chengjian

作者单位: Department of Electronic Engineering The University of Electronic Science Technology of China Chengdu, China

会议类型: 国际会议

会议名称: 2010 International Conference on Information Security and Artificial Intelligence(2010年信息安全与人工智能国际会议 ISAI 2010)

会议地点: 成都

会议语种:英文

页码: 481-483

在线出版日期: 2010-12-17（万方平台首次上网日期，不代表论文的发表时间）