/PERSON-RE-IDENTIFICATION-WITH-DEEP-DENSE-FEATURE-REPRESENTATION-AND-JOINT-BAYESIAN

Code for ICIP2017 accepted paper "PERSON RE-IDENTIFICATION WITH DEEP DENSE FEATURE REPRESENTATION AND JOINT BAYESIAN"

Primary LanguageMATLAB

PERSON-RE-IDENTIFICATION-WITH-DEEP-DENSE-FEATURE-REPRESENTATION-AND-JOINT-BAYESIAN

Code for ICIP2017 accepted paper "PERSON RE-IDENTIFICATION WITH DEEP DENSE FEATURE REPRESENTATION AND JOINT BAYESIAN"

"eval" contrains the evalution method "features" is the SD-CNN features

1.Some other prerequisites are

1.Matlab (to pre-process the CUHK03 dataset and Market1501 dataset)
2.python2 packages: numpy, scipy, Pillow, scikit-learn, protobuf

  1. run the scripts

    sudo chmod a+x ./ -R

    ./result_{dataset}.sh