Head Pose Classification Based on Line Portrait

Head pose classification has great significance in the field of biometrics. But most of the head pose classification methods depend on facial landmark detection, which results in that the performance relies not only on the pose classification. To this end, this paper proposes a head pose classification method based on the algorithm of line portrait generation and convolutional neural network (CNN). Experiments on the Color FERET and BIWI database show that the proposed algorithm achieves better results than the algorithm of head pose classification depend on facial landmark detection and has strong robustness.