(MGN_ReID) Multiple Granularity Network 多粒度行人重识别网络

Reproduction of paper: https://arxiv.org/abs/1804.01438 Learning Discriminative Features with Multiple Granularities for Person Re-Identification

Training Step

• Prepare data

  Donload: (Now support 4 datasets)

  • Market1501 data: http://www.liangzheng.org/Project/project_reid.html

  • DukeMTMC

  • Occ-Duke

  • MSMT17

  Put the data into ==/dataset== fold

• Modify the config.py

  • add dataset directoty to --root
  • change dataset name --dataset
  • modify other parameter

• Start to train

  run python train.py

• Test

  • After finishing the training, modify the --test_weight in config.py, then run python test.py

@ARTICLE{2018arXiv180401438W,
    author = {{Wang}, G. and {Yuan}, Y. and {Chen}, X. and {Li}, J. and {Zhou}, X.},
    title = "{Learning Discriminative Features with Multiple Granularities for Person Re-Identification}",
    journal = {ArXiv e-prints},
    archivePrefix = "arXiv",
    eprint = {1804.01438},
    primaryClass = "cs.CV",
    keywords = {Computer Science - Computer Vision and Pattern Recognition},
    year = 2018,
    month = apr,
    adsurl = {http://adsabs.harvard.edu/abs/2018arXiv180401438W},
    adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}

---

MGN论文复现: https://arxiv.org/abs/1804.01438 Learning Discriminative Features with Multiple Granularities for Person Re-Identification

训练步骤

• 数据准备

  下载: (目前支持4个数据集，后续会持续更新)

  • Market1501 data: http://www.liangzheng.org/Project/project_reid.html

  • DukeMTMC

  • Occ-Duke

  • MSMT17

  将数据放入 /dataset文件下

• 修改 config.py

  • 在 --root 中修改数据集位置
  • 更改数据集名字 --dataset （'market1501', 'occ_duke',......）
  • 更改其他参数

• 开始训练

  运行 python train.py

• 测试

  • 在训练完后 修改config.py中的--test_weight， 更改为得到的pth的地址, 然后运行 python test.py