Newborn face recognition using deep convolutional neural network

Development of expertise in Face Recognition has led researchers to apply its various techniques for newborn recognition as some of the problems such as swapping, kidnapping are still prevalent. The paper proposes to apply Deep Convolutional Neural Network(CNN) to IIT(BHU) newborn database. The database has its own advantages where the quality of images is high and segregation has been done for various expressions of newborn. The Deep CNN applied in this paper is more advantageous when compared to regular MLP. Along with this the results taken from application of proposed technique have been compared to state-of-the-art technique applied on the same database and it shows improved results. It has been found Deep CNN improves PCA by 22.09%, LDA by 12.98%, ICA by 11.35%, LBP by 17.08% and SURF by 10.8% for Neutral-Neutral faces. Along with this results have also been gathered to understand which Deep CNN architecture is most suitable for the database. The CNN architecture with 2 convolutional layers and 1 hidden layer is the best solution. The results have also been cross validated using 10-fold cross validation.

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

[2]  Mark D. McDonnell,et al.  Enhanced image classification with a fast-learning shallow convolutional neural network , 2015, 2015 International Joint Conference on Neural Networks (IJCNN).

[3]  Nitish Srivastava,et al.  Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[4]  Stefan Carlsson,et al.  CNN Features Off-the-Shelf: An Astounding Baseline for Recognition , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops.

[5]  Hari Om,et al.  (Two-Dimensional)2 whitening reconstruction for newborn recognition , 2017, Multimedia Tools and Applications.

[6]  Ah Chung Tsoi,et al.  Face recognition: a convolutional neural-network approach , 1997, IEEE Trans. Neural Networks.

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

[8]  Hari Om,et al.  Illumination Invariant Face Recognition of Newborn Using Single Gallery Image , 2016 .

[9]  Yoshua Bengio,et al.  Maxout Networks , 2013, ICML.

[10]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[11]  Tara N. Sainath,et al.  Improvements to Deep Convolutional Neural Networks for LVCSR , 2013, 2013 IEEE Workshop on Automatic Speech Recognition and Understanding.

[12]  Yann LeCun,et al.  What is the best multi-stage architecture for object recognition? , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[13]  Himanshu S. Bhatt,et al.  Domain Specific Learning for Newborn Face Recognition , 2016, IEEE Transactions on Information Forensics and Security.

[14]  Yoshua Bengio,et al.  Deep Sparse Rectifier Neural Networks , 2011, AISTATS.

[15]  Yaroslav Bulatov,et al.  Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks , 2013, ICLR.

[16]  Himanshu S. Bhatt,et al.  Face recognition for newborns: A preliminary study , 2010, 2010 Fourth IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS).

[17]  Hari Om,et al.  Pose Invariant Face Recognition for New Born: Machine Learning Approach , 2016 .

[18]  Sanjay Kumar Singh,et al.  Face recognition for newborns , 2012, IET Biom..

[19]  Ivan Laptev,et al.  Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[20]  Y. LeCun,et al.  Learning methods for generic object recognition with invariance to pose and lighting , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..