摘要
人脸关键点定位因受到表情、光照、姿态等的影响,常常会出现大的误差。为了准确地定位到人脸的关键点,提出了一种基于注意力模型的人脸关键点检测算法。先是利用可变型模型(DPM)算法检测出图片中的人脸区域,然后结合残差网络(ResNet)和收缩激励网络(SeNet)对该区域进行人脸关键点定位。实验结果表明,该算法在人脸数据集上获得了较高的准确率,证明了该算法的有效性。
Due to the influence of expressions, illuminations, gestures, etc., large errors often occur when positioning key points of a face. In order to accurately locate the key points of the face, a detection algorithm for key points on a face based on attention mechanism is proposed. Firstly, the deformable part model(DPM) algorithm is used to detect the face region in the picture, and then the focal point of the face is located in the region using ResNet and SeNet. The experimental results show that the algorithm achieves good accuracy on the face dataset, and prove the effectiveness of the algorithm.
作者
秦晓飞
盛凯
朱玥
杨勇
赵刚
贾程
李成名
鲁小东
周坚风
QIN Xiaofei;SHENG Kai;ZHU Yue;YANG Yong;ZHAO Gang;JIA Cheng;LI Chengming;LU Xiaodong;ZHOU Jianfeng(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Mechanical Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;Hangzhou Yimei Industrial Co.,Ltd.,Hangzhou 310000,China;Hangzhou Yimei Photoelectric Technology Co.,Ltd.,Hangzhou 310000,China)
出处
《光学仪器》
2020年第2期45-49,共5页
Optical Instruments
基金
国家重点研究发展计划(2016YFF0101400)。
