Edit the local of Caffe and dataset folder on file 1_, 2_, 6_, 7_...

Run these files in order 1_, 2_, 3_, 4_, 5_, 6_, 7_(optinal)...
1_Preparing training set and testing set
2_Making lmdb database
3_Making mean values
4_Convert mean. binaryproto to mean.py (option)
5_Fine-tuning or training
6_Testing or classifying (option)
7_Extracting feature (option)

CAVIAR4REID is a dataset for evaluating person re-identification algorithms. As the name suggest, the dataset has been extracted from the CAVIAR dataset mostly famous for person tracking and detection evaluations.

The results are generally around 80% on rank-1.