基于AdaBoost和显著性确认的 鲁棒人脸检测

In this paper, a robust face detection method using AdaBoost and saliency identification is proposed. The key point of this method is that saliency identification is used after AdaBoost-based face detection to eliminate fake faces with low saliency, and this helps to decrease the false positives in face detection. Experimental results demonstrate that this method can achieve lower false positives than AdaBoost method.