摘要: This paper deals with the use of confidence measure for Automatic Face Recognition (AFR). AFR is realized by adapted Kepenecki face recognition approach based on Gabor wavelet transform. work motivated fact that obtained rate real-world corpus only about 50% which not sufficient our application, a system automatic labelling photographs in large database. The main goal this thus proposition post-processing classification result order to remove “incorrectly” classified images. We show filter out incorrectly recognized faces beneficial. Two measures are proposed and evaluated Czech News Agency (CTK) corpus. Experimental results confirm benefit task.