Multi-View Gait Identification Based on Stacked Sparse Auto-Encoders

Due to the change of capture view angle, the intra-class variations of gait sequence get larger than inter-class variations in multi-view gait identification, which becomes one of the most intractable tasks. To overcome this problem effectively, a novel approach is proposed, which extracts the viewinvariance feature by a stacked sparse auto-encoders (SSAE) network. Each stacked layer maps a gait energy image (GEI) to the virtual one with smaller view angle variations. Repeat this process, a non-side GEI can be converted to a side one gradually. The feature variations caused by the change of view angle are extruded to zero step by step. The output of SSAE can be used as the discriminative feature for identity identification. Extensive evaluations are carried out based on the CASIA-B and OU-ISIR dataset. Results turn out that our method outperforms most state-of-the-art approaches, which shows great potential for practical application in future.

[1]  M. Lee,et al.  The University of Southampton Multi-Biometric Tunnel and introducing a novel 3D gait dataset , 2008, 2008 IEEE Second International Conference on Biometrics: Theory, Applications and Systems.

[2]  Qiang Wu,et al.  Multiple views gait recognition using View Transformation Model based on optimized Gait Energy Image , 2009, 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops.

[3]  Bir Bhanu,et al.  Individual recognition using gait energy image , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Z. Liu,et al.  Simplest representation yet for gait recognition: averaged silhouette , 2004, ICPR 2004.

[5]  Yongli Hu,et al.  Resampling based method for pixel-wise correspondence between 3D faces , 2004, International Conference on Information Technology: Coding and Computing, 2004. Proceedings. ITCC 2004..

[6]  Ahmed M. Elgammal,et al.  Gait tracking and recognition using person-dependent dynamic shape model , 2006, 7th International Conference on Automatic Face and Gesture Recognition (FGR06).

[7]  Yann LeCun,et al.  Convolutional networks and applications in vision , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[8]  Rama Chellappa,et al.  Towards a view invariant gait recognition algorithm , 2003, Proceedings of the IEEE Conference on Advanced Video and Signal Based Surveillance, 2003..

[9]  Qiang Wu,et al.  A New View-Invariant Feature for Cross-View Gait Recognition , 2013, IEEE Transactions on Information Forensics and Security.

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

[11]  Tao Xiang,et al.  Gait Recognition by Ranking , 2012, ECCV.

[12]  Yasushi Makihara,et al.  View Transformation Model Incorporating Quality Measures for Cross-View Gait Recognition , 2016, IEEE Transactions on Cybernetics.

[13]  Hua Li,et al.  3D gait recognition using multiple cameras , 2006, 7th International Conference on Automatic Face and Gesture Recognition (FGR06).

[14]  Jiwen Lu,et al.  Joint Subspace Learning for View-Invariant Gait Recognition , 2011, IEEE Signal Processing Letters.

[15]  Yasushi Makihara,et al.  Arbitrary view transformation model for gait person authentication , 2012, 2012 IEEE Fifth International Conference on Biometrics: Theory, Applications and Systems (BTAS).

[16]  Gunawan Ariyanto,et al.  Model-based 3D gait biometrics , 2011, 2011 International Joint Conference on Biometrics (IJCB).

[17]  Yap-Peng Tan,et al.  View invariant gait recognition , 2010, 2010 IEEE International Conference on Acoustics, Speech and Signal Processing.

[18]  Shaogang Gong,et al.  Cross View Gait Recognition Using Correlation Strength , 2010, BMVC.

[19]  Xiaogang Wang,et al.  A Comprehensive Study on Cross-View Gait Based Human Identification with Deep CNNs , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  Amit K. Roy-Chowdhury,et al.  A study on view-insensitive gait recognition , 2005, IEEE International Conference on Image Processing 2005.

[21]  Jiwen Lu,et al.  Uncorrelated discriminant simplex analysis for view-invariant gait signal computing , 2010, Pattern Recognit. Lett..

[22]  Dan Wang,et al.  Cross-View Gait Identification with Embedded Learning , 2017, ACM Multimedia.