抄録
H-034
Performance Improvement of PDLDA-based Face Recognition Using Fusion Face Descriptor
○I Gede Pasek Suta Wijaya（熊本大/Mataram University）・内村圭一・上瀧　剛（熊本大）
This paper presents multistage classifiers for improving the performance of PDLDA-based face recognition which can solve the large face variability due to pose variations. In this case, the PDLDA and fusion face descriptor are employed as the first and second stage classifier, respectively. The results show that the proposed method gives sufficient improvement of recognition rate against face pose variations when the tests were done in several challenges databases.