Improving Deep Unconstrained Facial Recognition by Data Augmentation

Facial recognition technology has emerged as an attractive solution for many of today’s needs in identification and identity verification. In recent years, the use of deep learning techniques and convolutional neural networks in particular has led to high-performance systems with near-human recognition capabilities. In general, these models are trained and evaluated on image datasets that do not sufficiently consider the lighting conditions of a real environment. However, in many practical applications the lighting is uncontrolled, which may seriously affect the performance of these systems. In this chapter, we propose a data augmentation method to achieve a model that is robust to variations in brightness. The training dataset is augmented by generating 3D faces from 2D images in the original dataset, followed by a Lambertian reflectance lighting variation that simulates the lighting variations that occur in real environments. The approach is evaluated on the YaleB and ORL datasets, with respective accuracy gains of 17.77% and 9%, compared to the model trained without data augmentation.

[1]  David J. Kriegman,et al.  From Few to Many: Illumination Cone Models for Face Recognition under Variable Lighting and Pose , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Ming Yang,et al.  DeepFace: Closing the Gap to Human-Level Performance in Face Verification , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[4]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[5]  Marwan Mattar,et al.  Labeled Faces in the Wild: A Database forStudying Face Recognition in Unconstrained Environments , 2008 .

[6]  John Lawrence,et al.  Comparing TensorFlow Deep Learning Performance Using CPUs, GPUs, Local PCs and Cloud , 2017 .

[7]  Habibollah Haron,et al.  Surface reconstruction techniques: a review , 2012, Artificial Intelligence Review.

[8]  James Philbin,et al.  FaceNet: A unified embedding for face recognition and clustering , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  Leon Sixt,et al.  RenderGAN: Generating Realistic Labeled Data , 2016, Front. Robot. AI.

[10]  Xiaogang Wang,et al.  Deep Learning Face Representation from Predicting 10,000 Classes , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[11]  Jun Wang,et al.  3D Facial Expression Recognition Based on Primitive Surface Feature Distribution , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[12]  Alexander M. Bronstein,et al.  Expression-Invariant 3D Face Recognition , 2003, AVBPA.

[13]  A. Bovik,et al.  Three Dimensional Face Recognition Using Iso-Geodesic and Iso-Depth Curves , 2008, 2008 IEEE Second International Conference on Biometrics: Theory, Applications and Systems.

[14]  Ronen Basri,et al.  Lambertian Reflectance and Linear Subspaces , 2003, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Thomas Vetter,et al.  A morphable model for the synthesis of 3D faces , 1999, SIGGRAPH.

[16]  David A. Landgrebe,et al.  A survey of decision tree classifier methodology , 1991, IEEE Trans. Syst. Man Cybern..

[17]  Sami Romdhani,et al.  A 3D Face Model for Pose and Illumination Invariant Face Recognition , 2009, 2009 Sixth IEEE International Conference on Advanced Video and Signal Based Surveillance.

[18]  Ying Wang,et al.  A survey on algorithms of hole filling in 3D surface reconstruction , 2016, The Visual Computer.

[19]  Marco Attene,et al.  Polygon mesh repairing: An application perspective , 2013, CSUR.

[20]  Kilian Q. Weinberger,et al.  Distance Metric Learning for Large Margin Nearest Neighbor Classification , 2005, NIPS.

[21]  Gui-Bin Bian,et al.  Performance Analysis of Google Colaboratory as a Tool for Accelerating Deep Learning Applications , 2018, IEEE Access.

[22]  Yiying Tong,et al.  Unconstrained 3D face reconstruction , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[24]  I. Jolliffe.  Principal Component Analysis , 2005 .

[25]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[26]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[27]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[28]  Andrew Zisserman,et al.  Deep Face Recognition , 2015, BMVC.

[29]  Davis E. King,et al.  Dlib-ml: A Machine Learning Toolkit , 2009, J. Mach. Learn. Res..

[30]  Gang Hua,et al.  Labeled Faces in the Wild: A Survey , 2016 .

[31]  Xiaogang Wang,et al.  DeepID3: Face Recognition with Very Deep Neural Networks , 2015, ArXiv.

[32]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[33]  Quoc V. Le,et al.  DropBlock: A regularization method for convolutional networks , 2018, NeurIPS.

[34]  Xiaogang Wang,et al.  Deeply learned face representations are sparse, selective, and robust , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[35]  A. Ganapathiraju,et al.  LINEAR DISCRIMINANT ANALYSIS - A BRIEF TUTORIAL , 1995 .

[36]  Gabriel Zachmann,et al.  Consistent normal orientation for polygonal meshes , 2004 .

[37]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[38]  Xi Zhou,et al.  Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network , 2018, ECCV.

[39]  Wen Gao,et al.  Efficient 3D reconstruction for face recognition , 2005, Pattern Recognit..

[40]  Xiaogang Wang,et al.  Deep Learning Face Representation by Joint Identification-Verification , 2014, NIPS.

[41]  Ahmet Çinar,et al.  Comparison of HOG (Histogram of Oriented Gradients) and Haar Cascade Algorithms with A Convolutional Neural Network Based Face Detection Approach , 2017 .

[42]  S. Shan,et al.  VIPLFaceNet: an open source deep face recognition SDK , 2016, Frontiers of Computer Science.

[43]  Ferdinando Silvestro Samaria,et al.  Face recognition using Hidden Markov Models , 1995 .