基于三维形变模型的人脸姿势表情校正

doi:10.11896/j.issn.1002-137X.2019.06.039

计算机科学 ›› 2019, Vol. 46 ›› Issue (6): 263-269.doi: 10.11896/j.issn.1002-137X.2019.06.039

所属专题：人脸识别

基于三维形变模型的人脸姿势表情校正

王钱庆¹, 张惊雷²

(天津理工大学天津市复杂系统控制理论及应用重点实验室天津300384)¹
(天津理工大学电气电子工程学院天津300384)²

收稿日期:2018-05-16 发布日期:2019-06-24
通讯作者: 张惊雷(1969-),男,博士,教授,主要研究方向为模式识别、图像的处理,E-mail:zhangjinglei@tjut.edu.cn
作者简介:王钱庆(1994-),女,硕士生,主要研究方向为图像处理、人脸三维重建等;

Face Pose and Expression Correction Based on 3D Morphable Model

WANG Qian-qing¹, ZHANG Jing-lei²

(Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems,Tianjin University of Technology,Tianjin 300384,China)¹
(School of Electrical and Electronics Engineering,Tianjin University of Technology,Tianjin 300384,China)²

Received:2018-05-16 Published:2019-06-24

摘要/Abstract

摘要： 针对目前人脸姿势校正鲁棒性差和计算复杂等问题,提出一种新的人脸姿态表情校正方法。首先,通过Fast-SIC算法来改进AAM模型以实现人脸对齐,该算法在不同光照、不同表情、不同姿势及不同遮挡的情况下均具有良好的对齐效果。然后,在人脸对齐的基础上进行人脸三维重建。文中提出了BFM-3DMM模型,其在原始3DMM模型的基础上添加了表情参数。但是,经过BFM-3DMM模型校正后的人脸仍然不够平滑,利用SFS算法不会受到原始统计模型约束的特点,对BFM-3DMM模型校正后的二维人脸进行再校正。在AFLW和LFPW数据库及自测人脸数据库上进行了相关实验,结果证明,校正后的二维人脸更加平滑且具有高保真度,还能够保留图像背景等信息。

关键词: 从阴影恢复形状法, 人脸三维重建, 三维形变模型, 主动表观模型

Abstract: Aiming at the problems such as poor robustness and computational complexity in face pose correction,a new facial pose and expression correction algorithmis was proposed.First,the Fast-SIC algorithm is adopted to improve the AAM model and to enhence the fitting efficiency.Then,based on the face alignment results,3D face reconstruction is performed.A BFM-3DMM model combining expression parameters into classical 3DMM model was proposed.However,the face corrected by the BFM-3DMM model is not smooth enough.Due to the fact that the SFS algorithm is not constrained by the original statistical model,this algorithm is applied to re-correct 2D face from BFM-3DMM model.The algorithm achieves good alignment and correction effects both on AFLW and LFPW,which are the two famous large face databases,as well as self-build face database.The experimental evaluation results show that the corrected 2D faces have smoother apperance and higher fidelity compared with classical 3DMM model,and can also retain image background information.

Key words: 3D Face reconstruction, 3D morphable model (3DMM), Active appearance model (AAM), Shape from shading algorithm (SFS)

中图分类号:

TP391

王钱庆, 张惊雷. 基于三维形变模型的人脸姿势表情校正[J]. 计算机科学, 2019, 46(6): 263-269. https://doi.org/10.11896/j.issn.1002-137X.2019.06.039

WANG Qian-qing, ZHANG Jing-lei. Face Pose and Expression Correction Based on 3D Morphable Model[J]. Computer Science, 2019, 46(6): 263-269. https://doi.org/10.11896/j.issn.1002-137X.2019.06.039

参考文献

[1]WANG J T,ZHAO L,QI X B.Face recognition method based on adaptive 3D Morphable Model and Multiple Manifold Discriminant Analysis[J].Computer Science,2017,44(S1):232-235.(in Chinese)
王渐韬,赵丽,齐兴斌.自适应三维形变模型结合流形分析的人脸识别方法[J].计算机科学,2017,44(S1):232-235.
[2]HUANG F,TAN S B.Face feature location based on improved active appearance model algorithm[J].Computer Engineering and Applications,2015,51(16):204-209.(in Chinese)
黄飞,谭守标.基于改进主动表观模型算法的人脸特征定位[J].计算机工程与应用,2015,51(16):204-209.
[3]TZIMIROPOULOS G,PANTIC M.Fast Algorithms for Fitting Active Appeaance Models to Unconstrained Images[J].International Journal of Computer Vision,2017,122(1):17-33.
[4]TRAN A T,HASSNER T,MASI I,et al.Regressing Robust and Discriminative 3D Morphable Models with a very Deep Neural Network[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).IEEE,2017:1493-1502.
[5]HU G,YAN F,KITTLER J,et al.Efficient 3D Morphable Face Model Fitting[J].Pattern Recognition,2017,67(C):366-379.
[6]GU Y R,QI R,WANG B Y.Research on 3D Face Attitude Correction Algorithm[J].Journal of Instrumentation and Instrument,2010,31(10):2291-2295.(in Chinese)
顾亦然,闵瑞,王保云.三维人脸姿态校正算法研究[J].仪器仪表学报,2010,31(10):2291-2295.
[7]FANG S Y,ZHOU D K,CAO Y P,et al.Positive Face Image Synthesis Based on Pose Estimation[J].Computer Engineering,2015,41(10):240-244.(in Chinese)
方三勇,周大可,曹元鹏,等.基于姿态估计的正面人脸图像合成[J].计算机工程,2015,41(10):240-244.
[8]DING L J,FENG H,HUANG Y.Geometric Algebra Invariant Method for 3D Face Attitude Correction[J].Microcomputer Systems,2015,36(1):177-181.(in Chinese)
丁立军,冯浩,黄宇.3D人脸姿态校正的几何代数不变量方法[J].小型微型计算机系统,2015,36(1):177-181.
[9]YANG B,MA L.Improvement of weak texture face image synthesis method based on deformation model[J].Computer Simulation,2016,33(9):248-251.(in Chinese)
杨勃,马禄.基于形变模型的弱纹理人脸图像合成方法改进[J].计算机仿真,2016,33(9):248-251.
[10]LUO Y,TAO Y,YANG G.Facial texture mapping and defor-mation based on facial feature constraints[J].Computer Engineering and Computer Engineering and Applications,2018,54(6):188-192,240.(in Chinese)
罗岱,陶洋,杨刚.基于面部特征约束的人脸纹理映射及变形[J].计算机工程与应用,2018,54(6):188-192,240.
[11]YIN X,XIANG Y,SOHN K,et al.Towards Large-Pose Face Frontalization in the Wild[C]∥International Conference on Computer Vision.2017.
[12]TZIMIROPOULOS G,PANTIC M.Optimization Problems for Fast AAM Fitting in-the-Wild[C]∥IEEE International Confe-rence on Computer Vision.IEEE Computer Society,2013:593-600.
[13]PAYSAN P,KNOTHE R,AMBERG B,et al.A 3d face model for pose and illumination invariant face recognition[C]∥Sixth IEEE International Conference on Advanced Video and Signal Based Surveillance,2009(AVSS’09).IEEE,2009:296-301.
[14]ZHU X,LEI Z,YAN J,et al.High-fidelity Pose and Expression Normalization for face recognition in thewild[C]∥Computer Vision and Pattern Recognition.IEEE,2015:787-796.
[15]CAO C,WENG Y,ZHOU S,et al.Faceware house:a 3d facial expression database for visual computing[J].IEEE Transactions on Visualization and Computer Graphics,2014,20(3):413-425.
[16]CHU B,ROMDHANI S,CHEN L.3D-Aided Face Recognition Robust to Expression and Pose Variations[C]∥IEEE Confe-rence on Computer Vision and Pattern Recognition.IEEE Computer Society,2014:1907-1914.
[17]BELHUMEUR D,JACOBS D,KRIEGMAN,et al.Localizing parts of faces using a consensus of exemplars[C]∥IEEE Conference on Computer Vision and Pattern Recognition.IEEE Computer Society,2011.
[18]KÖSTINGER M,WOHLHART P,ROTH P M,et al.Annotated facial landmarks in the wild:A large-scale,real-world database for facial landmark localization[C]∥2011 IEEE International Conference on Computer Vision Workshops (ICCV Workshops).IEEE,2011:2144-2151.
[19]ARASHLOO S R,KITTLER J.Efficient processing of MRFs for unconstrained-pose face recognition[C]∥IEEE Sixth International Conference on Biometrics:Theory,Applications and Systems.IEEE,2013:1-8.
[20]HASSNER T,HAREL S,PAZ E,et al.Effective face frontalization in unconstrained images[C]∥IEEE Conference on Compu-ter Vision and Pattern Recognition.IEEE Computer Society,2014:4295-4304.
[21]HUANG G B,MATTAR M A,LEE H,et al.Learning to align from scratch[C]∥International Conference on Neural Information Processing Systems.Curran Associates Inc.,2012:764-772.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于三维形变模型的人脸姿势表情校正

Face Pose and Expression Correction Based on 3D Morphable Model

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

Metrics

本文评价

推荐阅读 0