Person-Dependent and Person-Independent Facial Expression Recognition Based on 2DPCA Complementary Subspace
Xiaohua Chen · Journal of Huzhou Teachers College · 2009
Research shows that the information which is helpful to one recognition task would be interferential to another recognition task.Facial expression recognition calls for the information difference among all kinds of expressions.This paper proposes the calculation of expression recognition based on the two-dimensional principle analysis(2DPCA) complementary subspace according to the seven basic person-dependent and person-independent facial expressions.The results of the experiment made on the two JAFFE facial expression database of CED-WYU(1.0) and JAFFE show that the recognition rate by 2DPCA complementary subspace to determine person-dependent and person-independent expressions can reach 100%,which is higher than that by 2DPCA.