3D face recognition based on hierarchical feature network

doi:10.11772/j.issn.1001-9081.2020010103

Journal of Computer Applications ›› 2020, Vol. 40 ›› Issue (9): 2514-2518.DOI: 10.11772/j.issn.1001-9081.2020010103

• Artificial intelligence • Previous Articles Next Articles

3D face recognition based on hierarchical feature network

ZHAO Qing¹, YU Yuanhui²

1. School of Science, Jimei University, Xiamen Fujian 361021, China;
2. Computer Engineering College, Jimei University, Xiamen Fujian 361021, China

Received:2020-02-07 Revised:2020-05-31 Online:2020-09-10 Published:2020-06-03
Supported by:
This work is partially supported by Jimei University Horizontal Technology Development Project (S19142).

基于分层特征化网络的三维人脸识别

赵青¹, 余元辉²

1. 集美大学理学院, 福建厦门 361021;
2. 集美大学计算机工程学院, 福建厦门 361021

通讯作者: 余元辉
作者简介:赵青(1993-),女,河北辛集人,硕士研究生,主要研究方向:三维人脸识别;余元辉(1973-),男,河南平顶山人,副教授,硕士,主要研究方向:智能信息处理、移动agent计算、数据融合、多媒体图像处理。
基金资助:
集美大学横向技术开发项目（S19142）。

Abstract

Abstract: Focused on the problems of multiple expression variations, multiple pose variations as well as varying-degree missing face point cloud data in Three-Dimensional (3D) faces, 3D point cloud face data was exploratively applied to PointNet series classification networks, and the recognition results were compared and analyzed, then a new network framework named HFN (Hierarchical Feature Network) was proposed. First, the point cloud with fixed points was randomly sampled after data preprocessing. Second, the point fixed point cloud was input into SA (Set Abstraction) module in order to obtain the centroid points and neighborhood points of the local areas, and extract the features of the local areas, then the point cloud spatial structural features extracted from DSA (Directional Spatial Aggregation) module based on multi-directional convolution were mosaicked. Finally, the full connection layer was used to perform the classification of 3D faces, so as to realize the 3D face recognition. The results on CASIA database show that the average recognition rate of the proposed method is 96.34%, which is better than those of classification networks such as PointNet, PointNet++, PointCNN and Spatial Aggregation Net (SAN).

Key words: 3D face recognition, 3D point cloud, classification network, pose variation, spatial structural feature

摘要： 针对三维人脸多表情、多姿态变化同时存在，人脸点云数据不同程度缺失的问题，探索性地将三维点云人脸数据应用于PointNet系列的分类网络并进行了识别结果的对比与分析，然后提出了一种新的网络框架——HFN。首先，在数据预处理后随机采样固定点数的点云；其次，将固定点数的人脸点云输入SA模块，以获取局部区域的质心点、邻域点并提取局部区域的特征，然后拼接由DSA模块基于多方向卷积提取的点云空间结构特征；最后，利用全连接层进行三维人脸的分类，从而实现三维人脸识别。在CASIA数据库上的结果显示，所提方法的平均识别率为96.34%，优于PointNet、PointNet++、PointCNN和空间聚合网络（SAN）这几种分类网络。

关键词: 三维人脸识别, 三维点云, 分类网络, 姿态变化, 空间结构特征

CLC Number:

TP391.4

ZHAO Qing, YU Yuanhui. 3D face recognition based on hierarchical feature network[J]. Journal of Computer Applications, 2020, 40(9): 2514-2518.

赵青, 余元辉. 基于分层特征化网络的三维人脸识别[J]. 计算机应用, 2020, 40(9): 2514-2518.

References

[1] SOLTANPOUR S,BOUFAMA B,WU Q M J. A survey of local feature methods for 3D face recognition[J]. Pattern Recognition, 2017,72:391-406.
[2] DRIRA H,BEN AMOR B,SRIVASTAVA A,et al. 3D face recognition under expressions,occlusions,and pose variations[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013,35(9):2270-2283.
[3] LEI Y,BENNAMOUN M,HAYAT M,et al. An efficient 3D face recognition approach using local geometrical signatures[J]. Pattern Recognition,2014,47(2):509-524.
[4] XU C,LI S,TAN T,et al. Automatic 3D face recognition from depth and intensity Gabor features[J]. Pattern Recognition,2009, 42(9):1895-1905.
[5] SUN Y,WANG X,TANG X. Hybrid deep learning for face verification[C]//Proceedings of the 2013 IEEE International Conference on Computer Vision. Piscataway:IEEE,2013:1489-1496.
[6] KIM D, HERNANDEZ M, CHOI J, et al. Deep 3D face identification[C]//Proceedings of the 2017 IEEE International Joint Conference on Biometrics. Piscataway:IEEE,2017:133-142.
[7] LIU F,ZHAO Q,LIU X,et al. Joint face alignment and 3D face reconstruction with application to face recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2020, 42(3):664-678.
[8] GILANI S Z,MIAN A. Learning from millions of 3D scans for largescale 3D face recognition[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE,2018:1896-1905.
[9] MU G,HUANG D,HU G,et al. Led3D:a lightweight and efficient deep approach to recognizing low-quality 3D faces[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE,2019:5766-5775.
[10] QI C R,SU H,MO K,et al. PointNet:deep learning on point sets for 3d classification and segmentation[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE,2017:652-660.
[11] QI C R,YI L,SU H,et al. PointNet++:deep hierarchical feature learning on point sets in a metric space[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook,NY:Curran Associates Inc.,2017:5099-5108.
[12] LI Y,BU R,SUN M,et al. PointCNN:convolution on xtransformed points[C]//Proceedings of the 32nd International Conference on Neural Information Processing Systems. Cambridge:MIT Press,2018:820-830.
[13] CAI G,JIANG Z,WANG Z,et al. Spatial aggregation net:point cloud semantic segmentation based on multi-directional convolution[J]. Sensors,2019,19(19):No. 4329.
[14] JIANG M,WU Y,ZHAO T,et al. PointSIFT:a SIFT-like network module for 3D point cloud semantic segmentation[EB/OL].[2020-01-10]. https://arxiv.org/pdf/1807.00652.pdf.
[15] XU C,WANG Y,TAN T,et al. A robust method for detecting nose on 3D point cloud[J]. Pattern Recognition Letters,2006,27(13):1487-1497.
[16] Biometrics Ideal Test. CASIA-3D FaceV1[EB/OL].[2020-01-10]. http://biometrics.idealtest.org/dbDetailForUser.do?id=8.
[17] MIAN A,BENNAMOUN M,OWENS R. Automatic 3D face detection,normalization and recognition[C]//Proceedings of the 3rd International Symposium on 3D Data Processing, Visualization, and Transmission. Piscataway:IEEE, 2006:735-742.
[18] ELDAR Y,LINDENBAUM M,PORAT M,et al. The farthest point strategy for progressive image sampling[J]. IEEE Transactions on Image Processing,1997,6(9):1305-1315.
[19] 吴从中, 王浩宇, 詹曙. 融合曲面形状和纹理特征的三维人脸识别[J]. 电子测量与仪器学报,2018,32(9):150-156.(WU C Z,WANG H Y,ZHAN S. 3D face recognition by fusing surface shape and texture features[J]. Journal of Electronic Measurement and Instrumentation,2018,32(9):150-156.)

3D face recognition based on hierarchical feature network

基于分层特征化网络的三维人脸识别

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 2

Recommended Articles

Metrics

[1]	ZHAO Erping, LIU Wei, DANG Hong'en. Data compression and spatial indexing technology for massive 3D point cloud [J]. Journal of Computer Applications, 2018, 38(1): 146-151.
[2]	LU Le ZHOU Da-ke HU Yang-ming. 3D face reconstruction and recognition based on feature division [J]. Journal of Computer Applications, 2012, 32(11): 3189-3192.