Medina-Carnicer, R.Muñoz-Salinas, RafaelLópez-Fernández, D.Madrid-Cuevas, F.J.Carmona Poyato, Ángel2017-02-17T12:25:22Z2017-02-17T12:25:22Z2017-02-17http://hdl.handle.net/10396/14532Gait as biometrics has been widely used for
human identi cation. However, direction changes cause
di culties for most of the gait recognition systems, due
to appearance changes. This study presents an e cient
multi-view gait recognition method that allows curved
trajectories on completely unconstrained paths for in-
door environments. Our method is based on volumet-
ric reconstructions of humans, aligned along their way.
A new gait descriptor, termed as Gait Entropy Vol-
ume (GEnV), is also proposed. GEnV focuses on cap-
turing 3D dynamical information of walking humans
through the concept of entropy. Our approach does
not require the sequence to be split into gait cycles.
A GEnV based signature is computed on the basis of
the previous 3D gait volumes. Each signature is clas-
si ed by a Support Vector Machine, and a majority
voting policy is used to smooth and reinforce the clas-
si cations results. The proposed approach is experimen-
tally validated on the \AVA Multi-View Gait Dataset
(AVAMVG)" and on the \Kyushu University 4D Gait
Database (KY4D)". The results show that this new ap-
proach achieves promising results in the problem of gait
recognition on unconstrained paths.application/pdfenghttps://creativecommons.org/licenses/by-nc-nd/4.0/Gait entropy volumeGait recognitionView-Independent3D reconstructionEntropy Volumes for Viewpoint Independent Gait Recognitioninfo:eu-repo/semantics/preprintGobierno de España. TIN2012-32952info:eu-repo/semantics/openAccess