An automatic system for multi-view face detection and pose estimation

Over the past few years, multi-view face detection issue has become one of the most attractive research topics in the field of computer vision. In this paper, a novel automatic system for multi-view face detection and pose estimation is proposed. Our approach adopts modified appearance-based learning methods to build corresponding face detectors and pose estimators, and detects multi-view faces according to a coarse-to-fine structure. The experiments not only demonstrate the ability of our system to automatically identify facial images with a high degree of accuracy, but also verify its ability to estimate the pose angles.