A comparative study of Minimax Probability Machine-based approaches for face recognition |
| |
Affiliation: | 1. Graduate School at Shenzhen, Tsinghua University, Computer Science and Technology, Tsinghua Campus, The University Town, Shenzhen, Guangdong, China;2. Tsinghua University, Beijing, China |
| |
Abstract: | Automatic face recognition is a challenging problem in the biometric recognition area. Minimax Probability Machine (MPM) and its extension, Minimum Error Minimax Probability Machine, have shown advantages in the machine learning literature. In this paper, we incorporate the MPM-based approaches into our face recognition system for further study. To test the performance of our new system, we compare the MPM-based approaches with SVM, a PCA-based and a LDA-based algorithms on the FERET database for both verification and identification. The experimental results demonstrate that MPM-based approaches are promising for automatic face recognition. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|