光电工程, 2020, 47 (1): 190299, 网络出版: 2020-02-24  

基于深度学习检测器的多角度人脸关键点检测

Multi-angle key point detection of face based on deep learning detector
作者单位
杭州电子科技大学自动化学院,浙江 杭州 310018
摘要
针对人脸关键点检测(人脸对齐)在应用场景下的速度和精度需求, 首先在 SSD基础之上融合更多分布均匀的特征层, 对人脸框坐标进行级联预测, 形成对于多尺度人脸信息均具有更加鲁棒响应的深度学习检测器 MR-SSD。其次在局部二值特征 LBF的级联形状回归方法基础上, 提出了基于面部像素差值的多角度初始化算法。采用端正人脸正负 90°倾斜范围内的五组特征点形状进行初始化, 求取每组回归后形状的眼部特征点像素均方差值并以最大者对应方案作为最终回归形状, 从而实现对多角度倾斜人脸优异的拟合效果。本文所提出的最优架构可以实时获得极具鲁棒性的人脸框坐标并且可实现对于多角度倾斜人脸的关键点检测。
Abstract
In order to meet the speed and accuracy requirements of face key point detection (face alignment) in ap-plication scenarios, firstly, cascaded prediction is carried out on the basis of SSD (single shot multibox detector),which combines more uniformly distributed feature layers to form MR-SSD (more robust SSD), a deep learning de-tector with more robust response to multi-scale faces. Secondly, based on the cascade shape regression method oflocal binary feature (LBF), a multi-angle initialization algorithm based on the difference between the facial pixels isproposed. Five groups of feature points in the 90 degree inclination range of positive and negative face are initializedto achieve excellent fitting effect for inclined face under multi angles. The mean square deviation of each group offeature points after regression is calculated and the maximum corresponding shape is used as the final regressionshape. The optimal architecture proposed in this paper can obtain robust face bounding box and face alignmentschemes against multi-angle tilt in real time.

赵兴文, 杭丽君, 宫恩来, 叶锋, 丁明旭. 基于深度学习检测器的多角度人脸关键点检测[J]. 光电工程, 2020, 47(1): 190299. Zhao Xingwen, Hang Lijun, Gong Enlai, Ye Feng, Ding Mingxu. Multi-angle key point detection of face based on deep learning detector[J]. Opto-Electronic Engineering, 2020, 47(1): 190299.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!