/2016_person_re-ID

A Discriminatively Learned CNN Embedding for Person Re-identification

Primary LanguageHTMLOtherNOASSERTION

A Discriminatively Learned CNN Embedding for Person Re-identification

In this package, we provide our training and testing code written in Matconvnet for the paper [A Discriminatively Learned CNN Embedding for Person Re-identification] (https://arxiv.org/abs/1611.05666).

We also include matconvnet-beta23 which has been modified for our paper. All codes have been test on Ubuntu14.04 and Ubuntu16.04 with Matlab R2015b.

#Dataset Download [Market1501 Dataset] (http://www.liangzheng.org/Project/project_reid.html) #Models I use the git-lfs to store the large models. But this reposity may be over data quota. Alternatively, you can download the complete codes from GoogleDriver or [BaiduYun] (https://pan.baidu.com/s/1mhKoQ4S). BaiduYun sometime changes the link. If you find the url fail, you can contact with me to update it.

#To Test

  1. Use start-zzd.sh to start matlab. (You need to add your CUDA path in it. Then just type ./start-zzd.sh to run it.)

  2. Compile matconvnet. (You just need to uncomment and modify some lines in gpu_compile.m and run it. Try it~) If you fail in compilation, you may refer to http://www.vlfeat.org/matconvnet/install/

  3. After compilation, run test2/test_gallery_res.m to extract the features of gallery. They will store in a .mat file. You can use it to do evaluation. (For example, you may modify the Market1501 baseline code to evaluate our model. It may take a while.)

#To Train

  1. Compile matconvnet. If you fail in compilation, you can refer to http://www.vlfeat.org/matconvnet/install/

  2. Add your dataset path into prepare_data.m and run it. Make sure the code outputs the right image path.

  3. Run train_id_net_res_2stream.m to have fun.

Citation

Please cite this paper in your publications if it helps your research:

@article{zheng2016discriminatively,
  title={A Discriminatively Learned CNN Embedding for Person Re-identification},
  author={Zheng, Zhedong and Zheng, Liang and Yang, Yi},
  journal={arXiv preprint arXiv:1611.05666},
  year={2016}
}

#Thanks Thanks for Xuanyi Dong to realize our paper in Caffe.

Thanks for Weihang Chen to report the bug in prepare_data.m.