A Novel Survey Based on Multiethnic Facial Semantic Web

Abstract

The face includes a number of facial features which are various in minorities. Firstly, according to the correlations of the face parts shape semantics, multiethnic facial semantic web is proposed. It represents the relationship which belongs to the same minority and the difference of that belongs to the different minorities. Secondly, multiethnic facial semantic web is reduced by the correlations between the parts of the face. The semantic web which is reduced can maintains most available information which is belong to original semantic web, reduces the complexity and indirectly analysis the national facial features. Lastly, the effectiveness of our experiment is demonstrated by some real-word data sets.

DOI: http://dx.doi.org/10.11591/telkomnika.v11i9.3251