Department of Computing and Numerical Analysis, University of Cordoba
Department of Computer Architecture, University of Málaga
Department of Computer Science and Artificial Intelligence, University of Granada
 UCO

CNN for gait recognition

Francisco M. Castro, Manuel J. Marín-Jiménez, Nicolás Guil, Nicolás Pérez de la Blanca

THIS IS CURRENTLY A PRIVATE URL. PLEASE, DON'T SHARE

Overview

CNN filters We release here pretrained models [1] for gait signature extraction from: gray pixels, optical flow channels and depth maps. These models are ready to be used with MatConvNet library on TUM-GAID and CASIA-B datasets.


Downloads

Filename Description Size
gray-tum.zip CNN model for gray pixels - TUM-GAID 140 MB
of-tum.zip CNN model for optical flow - TUM-GAID 141 MB
depth-tum.zip CNN model for depth maps - TUM-GAID 140 MB
gray-casiab.zip CNN model for gray pixels - CASIA-B 143 MB
of-casiab.zip CNN model for optical flow - CASIA-B 138 MB
drawgaitfilters.zip Matlab code to display the convolutional filters stored in the CNN models. 2 KB

Related Publications

[1] F.M. Castro, M. Marin-Jimenez, N. Guil, N. Perez de la Blanca
Multimodal CNN for people identification
Under review, 2017

Acknowledgements

This work has been partially funded by the Research Project TIC-1692 (Junta de Andalucía).