Abstract: To deal with the problem of the traditional multi-pose face recognition based on the multi-view that it must have many face photos, a multi-pose face recognition algorithm based on the two orthogonal views is proposed. It consists of two steps. In the first step, a 3D face model is built by the person's two orthogonal views and multi-pose face images are produced by the projections of the 3D face model. In the second step, multi-pose face recognition is performed based on the training set that consists of the two orthogonal views and the synthesized multi-pose images. The experimental results show that the performance of the algorithm discussed in the paper is by far superior to that of the method based on the front view.

Key words: face recognition, multi-pose, orthogonal view, 3D model