A novel approach of gait recognition through fusion with footstep information

This paper is focused on two biometric modes which are very linked together: gait and footstep biometrics. Footstep recognition is a relatively new biometric based on signals extracted from floor sensors, while gait has been more researched and it is based on video sequences of people walking. This paper reports a directly comparative assessment of both biometrics using the same database (SFootBD) and experimental protocols. A fusion of the two modes leads to an enhanced gait recognition performance, as the information from both modes comes from different capturing devices and is not very correlated. This fusion could find application in indoor scenarios where a gait recognition system is present, such as in security access (e.g. security gate at airports) or smart homes. Gait and footstep systems achieve results of 8.4% and 10.7% EER respectively, which can be significantly improved to 4.8% EER with their fusion at the score level into a walking biometric.

[1]  F. Livesey,et al.  The ORL Acfive Floor , 1997 .

[2]  F. Livesey,et al.  The ORL active floor [sensor system] , 1997, IEEE Wirel. Commun..

[3]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[4]  Juha Röning,et al.  Methods for person identification on a pressure-sensitive floor: Experiments with multiple classifiers and reject option , 2008, Inf. Fusion.

[5]  Mark S. Nixon,et al.  What image information is important in silhouette-based gait recognition? , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[6]  Haiping Lu,et al.  Multilinear Principal Component Analysis of Tensor Objects for Recognition , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[7]  Juha Röning,et al.  Combining classifiers with different footstep feature sets and multiple samples for person identification , 2005, Proceedings. (ICASSP '05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005..

[8]  Wei Xiong,et al.  Active energy image plus 2DLPP for gait recognition , 2010, Signal Process..

[9]  Jeha Ryu,et al.  The User Identification System Using Walking Pattern over the ubiFloor , 2003 .

[10]  Gonzalo Bailador,et al.  Application of the computational theory of perceptions to human gait pattern recognition , 2010, Pattern Recognit..

[11]  Mark S. Nixon,et al.  Automatic Recognition by Gait , 2006, Proceedings of the IEEE.

[12]  Walid Karam,et al.  Talking-Face Identity Verification, Audiovisual Forgery, and Robustness Issues , 2009, EURASIP J. Adv. Signal Process..

[13]  Brian R. Mace,et al.  Person recognition by measuring the ground reaction force due to a footstep , 1970 .

[14]  Mark S. Nixon,et al.  A floor sensor system for gait recognition , 2005, Fourth IEEE Workshop on Automatic Identification Advanced Technologies (AutoID'05).

[15]  Gregory D. Abowd,et al.  The smart floor: a mechanism for natural user identification and tracking , 2000, CHI Extended Abstracts.

[16]  Philippe C. Cattin,et al.  Biometric authentication system using human gait , 2002 .

[17]  Jie Yang,et al.  Gait recognition based on dynamic region analysis , 2008, Signal Process..

[18]  M. D. Addlesee,et al.  The ORL Active Floor , 1997 .

[19]  Julian Fiérrez,et al.  Comparative Analysis and Fusion of Spatiotemporal Information for Footstep Recognition , 2013, IEEE Transactions on Pattern Analysis and Machine Intelligence.

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

[21]  Julian Fierrez,et al.  Analysis of spatial domain information for footstep recognition , 2011 .