The faces in video sequences were predicted via motion quotiety and horizontal and vertical adjustment factors.

• 运用运动系数加上横向和纵向调节因子，对后续帧中的人脸加以预测。