首页 > 论文 > 液晶与显示 > 33卷 > 3期(pp:254-260)

基于SURF特征的栈式自编码网络人脸对齐算法

Stacked auto-encoder networks face alignment algorithm based on SURF features

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

人脸对齐是人脸识别系统中的一个核心部分, 定位的准确性和定位速度直接影响到人脸识别的效果。人脸图像存在不同姿态、不同表情、不同光照条件等因素的影响, 真实场景下的人脸对齐成为一个难题。本文提出了一种基于SURF特征的栈式自编码网络人脸对齐方法, 首先通过粗糙定位网络找到近似人脸特征点, 并提取局部的SURF特征, 输入到局部细化网络, 通过级联结构, 进一步精确化人脸特征点的具体位置。最后, 在人脸数据集AFLW和HELEN上与近几年的对齐方法进行对比实验, 平均错误率8.80%, i5四核CPU, 2.3 Hz主频硬件平台下计算时间76 ms。我们的人脸对齐方法在真实场景下(包括单人和多人)具有较好的鲁棒性, 可以实现准确定位。

Abstract

Face alignment is an important part of face recognition system. The accuracy and speed of alignment directly affect the effect of face recognition. Face images have different postures, different expressions, different lighting conditions and other factors. Face alignment becomes a difficult problem. This paper proposes a auto-encoder network face alignment algorithm based on SURF features. The approximate face feature points are found by the rough location network. The local SURF features are extracted and input into the local thinning network. The exact locations of the feature points of the face are further refined by the cascade structure. In the face data sets AFLW and HELEN, we compared some new alignment method. The average error rate is 8.80%, the computation time is 7.6 ms under i5 quad core CPU, 2.3 Hz dominant frequency. Our face alignment method has good robustness in real scene (including single and multiple faces), and it can achieve accurate positioning.

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP391.41

DOI:10.3788/yjyxs20183303.0254

所属栏目:图像处理

基金项目:国家自然科学基金(No.61172111); 吉林省科技发展计划(No.20160101260JC); 吉林省教育厅“十三五”科学技术项目(No.JJKH20170625KJ)

收稿日期:2017-11-27

修改稿日期:2018-01-05

网络出版日期:--

作者单位    点击查看

崔 凯:长春理工大学 电子信息工程学院, 吉林 长春 130022
才 华:长春理工大学 电子信息工程学院, 吉林 长春 130022
刘广文:长春理工大学 电子信息工程学院, 吉林 长春 130022
刘 智:长春理工大学 电子信息工程学院, 吉林 长春 130022

联系人作者:崔凯(944515680@qq.com)

备注:崔凯(1993-), 男, 山东潍坊人, 硕士研究生, 主要研究领域为机器学习和机器视觉。

【1】ZHU X X, RAMANAN D. Face detection, pose estimation, and landmark localization in the wild [C]//Proceeding of the IEEE Conference on Computer Vision and Pattern Recognition. Providence, RI, USA: IEEE, 2012: 2879-2886.

【2】宿佳宁.人脸对齐算法研究 [D].哈尔滨: 哈尔滨工程大学, 2012.
SU J N. Research on face alignment algorithm [D]. Harbin: Harbin Engineering University, 2012. (in Chinese)

【3】BEHAINE C A R, SCHARCANSKI J. Enhancing the performance of active shape models in face recognition applications [J]. IEEE Transactions on Instrumentation and Measurement, 2012, 61(8): 2330-2333.

【4】COOTES T F, EDWARDS G J, TAYLOR C J. Active appearance models [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(6): 681-685.

【5】YU X, HUANG J Z, ZHANG S T, et al. Pose-free facial landmark fitting via optimized part mixtures and cascaded deformable shape model [C]//Proceeding of the IEEE International Conference on Computer Vision. Sydney, NSW, Australia: IEEE, 2013: 1944-1951.

【6】CRISTINACCE D, COOTES T. Automatic feature localisation with constrained local models [J]. Pattern Recognition, 2008, 41(10): 3054-3067.

【7】XIONG X H, DE LA TORRE F. Supervised descent method and its applications to face alignment [C]//Proceedings of 2013 IEEE Conference on Computer Vision and Pattern Recognition. Portland, OR, USA: IEEE, 2013: 532-539.

【8】刘洋, 韩广良, 史春蕾.基于SIFT算法的多表情人脸识别 [J].液晶与显示, 2016, 31(12): 1156-1160.
LIU Y, HAN G L, SHI C L. Recognition of expression-variant faces based on SIFT method [J]. Chinese Journal of Liquid Crystals and Displays, 2016, 31(12): 1156-1160. (in Chinese)

【9】CUI Z, LI W, XU D, et al. Fusing robust face region descriptors via multiple metric learning for face recognition in the wild [C]//Proceeding of 2013 IEEE Conference on Computer Vision and Pattern Recognition. Portland, OR, USA: IEEE, 2013: 3554-3561.

【10】LAGANIERE R.OpenCV 2计算机视觉编程手册 [M].张静, 译.北京: 科学出版社, 2013: 185-188.
LAGANIERE R. OpenCV 2 Computer Vision Programming Manual [M]. Zhang J, trans. Beijing: Science Press, 2013: 185-188. (in Chinese)

【11】许佳佳.结合Harris与SIFT算子的图像快速配准算法 [J].中国光学, 2015, 8(4): 574-581.
XU JJ. Fast image registration method based on Harris and SIFT algorithm [J]. Chinese Optics, 2015, 8(4): 574-581. (in Chinese)

【12】ZHANG J, SHAN S G, KAN M N, et al. Coarse-to-Fine Auto-Encoder Networks (CFAN) for real-time face alignment [C]//Proceeding of the 13th European Conference on Computer Vision. Zurich, Switzerland: Springer, 2014: 1-16.

【13】REN S Q, CAO X D, WEI Y C, et al. Face alignment at 3000 FPS via regressing local binary features [C]//Proceedings of the Conference on Computer Vision and Pattern Recognition. Columbus, OH, USA: IEEE, 2014: 1685-1692.

【14】姚文韬, 沈春锋, 顾志松, 等.基于多尺度SDM模型的人脸对齐算法 [J].控制工程, 2015, 22(S): 110-114.
YAO W T, SHEN C F, GU Z S, et al. Face alignment based on multi-scale SDM model [J]. Control Engineering of China, 2015, 22(S): 110-114. (in Chinese)

【15】彭明超, 包姣, 叶茂, 等.基于形状参数回归的人脸对齐算法 [J].模式识别与人工智能, 2016, 29(1): 63-71.
PENG M C, BAO J, YE M, et al. Face alignment algorithm based on shape parameter regression [J]. Pattern Recognition and Artificial Intelligence, 2016, 29(1): 63-71. (in Chinese)

引用该论文

CUI Kai,CAI Hua,LIU Guang-wen,LIU Zhi. Stacked auto-encoder networks face alignment algorithm based on SURF features[J]. Chinese Journal of Liquid Crystals and Displays, 2018, 33(3): 254-260

崔 凯,才 华,刘广文,刘 智. 基于SURF特征的栈式自编码网络人脸对齐算法[J]. 液晶与显示, 2018, 33(3): 254-260

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF