Time filter

Source Type

Chen M.,Heilongjiang University | Wang C.,Zhengzhou Kindergartens Teachers College | Meng X.,Heilongjiang University | Wang Z.,Heilongjiang University
Advances in Intelligent Systems and Computing | Year: 2015

This paper proposes a multi-modal fusion method that based on quaternion, and principal component analysis (PCA) in quaternion field is involved in our algorithm. We can fuse four different features into quaternion and complete the recognition process in quaternion field. This algorithm reduces the equal error rate (EER) while fusing more kinds of features. Our experiments that fuses three kinds of modalities and four different features with two kinds of modalities respectively show a observably improvement on recognition rate with the proposed algorithm. © Springer International Publishing Switzerland 2016. Source

Discover hidden collaborations