An Efficient Method for Face Localization and Recognition in Color Images

Face Localization Using an Effective Co-evolutionary Genetic Algorithm

“…The outcome of experimental results is shown in the Table II. This table shows that the third category (n=9,…,12) is more appropriate for classification of faces within the proposed system. To compare this system with the system presented in [10], the classification error and classification time of the two systems are listed in Table III. These experiments have been carried out by using a Pentium 4 processor, in the Matlab (version 7.2) environment.…”

“…For this purpose, we use an RBF neural network with the structure presented in [10] for classification of feature vectors. This structure is depicted in Fig.…”

“…The number of nodes in the output layer is set equal to the number of face classes. In this structure, the RBF unit parameters and output connection weights are determined using the procedures presented in [10].…”

“…Then, by using the location of eyes in the face image, the parameters of face's bounding ellipse (center, orientation, major and minor axis) are computed and the face's ellipse is extracted. Next, for evaluating of the proposed system we use Pseudo Zernike Moments (PZM) to generate the feature vector elements and an RBF neural network with the structure described in [10] as the classifier. Experiments show that the accuracy and speed of the proposed method are higher than the system proposed in [10].…”

Two faces are better than one: Face recognition in group photographs

“…Such systems usually have a pre-processing stage where color and lighting intensity variations are normalized before being discarded. On the other hand, there have been many attempts which show improvements in accuracy when color is leveraged for recognition [22][23][24].…”