Pytorch ReID

Pytorch ReID

Strong, Small, Friendly

Python3.6+ License: MIT

A tiny, friendly, strong baseline code for Object-reID (based on pytorch) since 2017.

Code is at .

Share to Facebook Twitter Weibo LinkedIn Email


Table of contents


Now we have supported:


  • Running the code on Google Colab with Free GPU. Check Here (Thanks to @ronghao233)
  • DG-Market (10x Large Synethic Dataset from Market CVPR 2019 Oral)
  • Swin Transformer / EfficientNet / HRNet
  • ResNet/ResNet-ibn/DenseNet
  • Circle Loss, Triplet Loss, Contrastive Loss, Sphere Loss, Lifted Loss, Arcface, Cosface and Instance Loss
  • Float16 to save GPU memory based on apex
  • Part-based Convolutional Baseline(PCB)
  • Random Erasing
  • Linear Warm-up


  • TensorRT
  • Pytorch JIT
  • Fuse Conv and BN layer into one Conv layer
  • Multiple Query Evaluation
  • Re-Ranking (CPU & GPU Version)
  • Visualize Training Curves
  • Visualize Ranking Result
  • Visualize Heatmap

Here we provide hyperparameters and architectures, that were used to generate the result. Some of them (i.e. learning rate) are far from optimal. Do not hesitate to change them and see the effect.

P.S. With similar structure, we arrived Rank@1=87.74% mAP=69.46% with Matconvnet. (batchsize=8, dropout=0.75) You may refer to Here. Different framework need to be tuned in a different way.

Some News

12 Aug 2023 Large Person Langauge Model is currently available at HereGitHub stars accepted by ACM MM’23. You are welcomed to check it.

19 Mar 2023 We host a special session on IEEE Intelligent Transportation Systems Conference (ITSC), covering the object re-identification & point cloud topic. The paper ddl is by May 15, 2023 and the paper notification is at June 30, 2023. Please select the session code ``w7r4a’’ during submission. More details can be found at Special Session Website.

9 Mar 2023 Market-1501 is in 3D. Please check our single 2D to 3D reconstruction work GitHub stars. 3D Human Animation

2022 News **7 Sep 2022** We support SwinV2. **24 Jul 2022** Market-HQ is released with super-resolution quality from 128\*64 to 512\*256. Please check at **14 Jul 2022** Add adversarial training by ``python --name ftnet_adv --adv 0.1 --aiter 40``. **1 Feb 2022** Speed up the inference process about 10 seconds by removing the ``cat`` function in ````. **1 Feb 2022** Add the demo with ``TensorRT`` (The fast inference speed may depend on the GPU with the latest RT Core).
2021 News **30 Dec 2021** We add supports for new losses, including arcface loss, cosface loss and instance loss. The hyper-parameters are still tunning. **3 Dec 2021** We add supports for four losses, including triplet loss, contrastive loss, sphere loss and lifted loss. The hyper-parameters are still tunning. **1 Dec 2021** We support EfficientNet/HRNet. **15 Sep 2021** We support ResNet-ibn from ECCV2018 ( **17 Aug 2021** We support running code on Google Colab with free GPU. Please check it out at . **14 Aug 2021** We have supported the training with [DG-Market]( for regularization via [Self-supervised Memory Learning]( You only neeed to download/unzip the dataset and add `--DG` to train model. **12 Aug 2021** We have supported the transformer-based model `Swin` by `--use_swin`. The basic performance is 92.73% Rank@1 and 79.71%mAP. **23 Jun 2021** Attack your re-ID model via Query! They are not robust as you expected! Check the code at [Here]( **5 Feb 2021** We have supported [Circle loss]( Oral). You can try it by simply adding `--circle`. **11 January 2021** On the Market-1501 dataset, we accelerate the re-ranking processing from **89.2s** to **9.4ms** with one K40m GPU, facilitating the real-time post-processing. The pytorch implementation can be found in [GPU-Re-Ranking](GPU-Re-Ranking/).
2020 News **11 June 2020** People live in the 3D world. We release one new person re-id code [Person Re-identification in the 3D Space](, which conduct representation learning in the 3D space. You are welcomed to check out it. **30 April 2020** We have applied this code to the [AICity Challenge 2020](, yielding the 1st Place Submission to the re-id track :red_car:. Check out [here]( **01 March 2020** We release one new image retrieval dataset, called [University-1652](, for drone-view target localization and drone navigation :helicopter:. It has a similar setting with the person re-ID. You are welcomed to check out it.
2019 News **07 July 2019:** I added some new functions, such as `--resume`, auto-augmentation policy, acos loss, into [developing thread]( and rewrite the `save` and `load` functions. I haven't tested the functions throughly. Some new functions are worthy of having a try. If you are first to this repo, I suggest you stay with the master thread. **01 July 2019:** [My CVPR19 Paper]( is online. It is based on this baseline repo as teacher model to provide pseudo label for the generated images to train a better student model. You are welcomed to check out the opensource code at [here]( **03 Jun 2019:** Testing with multiple-scale inputs is added. You can use `--ms 1,0.9` when extracting the feature. It could slightly improve the final result. **20 May 2019:** Linear Warm Up is added. You also can set warm-up the first K epoch by `--warm_epoch K`. If K <=0, there will be no warm-up.
2018 & 2017 News **What's new:** FP16 has been added. It can be used by simply added `--fp16`. You need to install [apex]( and update your pytorch to 1.0. Float16 could save about 50% GPU memory usage without accuracy drop. **Our baseline could be trained with only 2GB GPU memory.** ```bash python --fp16 ``` **What's new:** Visualizing ranking result is added. ```bash python python python python --query_index 777 ``` **What's new:** Multiple-query Evaluation is added. The multiple-query result is about **Rank@1=91.95% mAP=78.06%**. ```bash python python python --multi python ``` **What's new:**  [PCB]( is added. You may use '--PCB' to use this model. It can achieve around **Rank@1=92.73% mAP=78.16%**. I used a GPU (P40) with 24GB Memory. You may try apply smaller batchsize and choose the smaller learning rate (for stability) to run. (For example, `--batchsize 32 --lr 0.01 --PCB`) ```bash python --PCB --batchsize 64 --name PCB-64 python --PCB --name PCB-64 ``` **What's new:** You may try `` to conduct a faster evaluation with GPU. **What's new:** You may apply '--use_dense' to use `DenseNet-121`. It can arrive around Rank@1=89.91% mAP=73.58%. **What's new:** Re-ranking is added to evaluation. The re-ranked result is about **Rank@1=90.20% mAP=84.76%**. **What's new:** Random Erasing is added to train. **What's new:** I add some code to generate training curves. The figure will be saved into the model folder when training. ![](

Trained Model

I re-trained several models, and the results may be different with the original one. Just for a quick reference, you may directly use these models. The download link is Here.

Methods Rank@1 mAP Reference
[EfficientNet-b4] 85.78% 66.80% python --use_efficient --name eff; python --name eff
[ResNet-50 + adv defense] 87.77% 69.83% python --name adv0.1_40_w10_all --adv 0.1 --aiter 40 --warm 10 --train_all; python --name adv0.1_40_w10_all
[ConvNeXt] 88.98% 71.35% python --use_convnext --name convnext; python --name convnext
[ResNet-50 (fp16)] 88.03% 71.40% python --name fp16 --fp16 --train_all
[ResNet-50] 88.84% 71.59% python --train_all
[ResNet-50-ibn] 89.13% 73.40% python --train_all --name res-ibn --ibn
[DenseNet-121] 90.17% 74.02% python --name ft_net_dense --use_dense --train_all
[DenseNet-121 (Circle)] 91.00% 76.54% python --name ft_net_dense_circle_w5 --circle --use_dense --train_all --warm_epoch 5
[HRNet-18] 90.83% 76.65% python --use_hr --name hr18; python --name hr18
[PCB] 92.64% 77.47% python --name PCB --PCB --train_all --lr 0.02
[PCB + DG] 92.70% 78.31% python --name PCB_DG --PCB --train_all --lr 0.02 --DG; python --name PCB_DG
[ResNet-50 (all tricks)] 91.83% 78.32% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 8 --lr 0.02 --name warm5_s1_b8_lr2_p0.5
[ResNet-50 (all tricks+Circle)] 92.13% 79.84% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 8 --lr 0.02 --name warm5_s1_b8_lr2_p0.5_circle --circle
[ResNet-50 (all tricks+Circle+DG)] 92.13% 80.13% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 8 --lr 0.02 --name warm5_s1_b8_lr2_p0.5_circle_DG --circle --DG; python --name warm5_s1_b8_lr2_p0.5_circle_DG
[DenseNet-121 (all tricks+Circle)] 92.61% 80.24% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 8 --lr 0.02 --name dense_warm5_s1_b8_lr2_p0.5_circle --circle --use_dense; python --name dense_warm5_s1_b8_lr2_p0.5_circle
[HRNet-18 (all tricks+Circle+DG)] 92.19% 81.00% python --use_hr --name hr18_p0.5_circle_w5_b16_lr0.01_DG --lr 0.01 --batch 16 --DG --erasing_p 0.5 --circle --warm_epoch 5; python --name hr18_p0.5_circle_w5_b16_lr0.01_DG
[Swin] (224x224) 92.75% 79.70% python --use_swin --name swin; python --name swin
[SwinV2 (all tricks+Circle 256x128)] 92.93% 82.99% python --use_swinv2 --name swinv2_p0.5_circle_w5_b16_lr0.03 --lr 0.03 --batch 16 --erasing_p 0.5 --circle --warm_epoch 5; python --name swinv2_p0.5_circle_w5_b16_lr0.03 --batch 32
[Swin (all tricks+Circle 224x224)] 94.12% 84.39% python --use_swin --name swin_p0.5_circle_w5 --erasing_p 0.5 --circle --warm_epoch 5; python --name swin_p0.5_circle_w5
[Swin (all tricks+Circle+b16 224x224)] 94.00% 85.21% python --use_swin --name swin_p0.5_circle_w5_b16_lr0.01 --lr 0.01 --batch 16 --erasing_p 0.5 --circle --warm_epoch 5; python --name swin_p0.5_circle_w5_b16_lr0.01
[Swin (all tricks+Circle+b16+DG 224x224)] 94.00% 85.36% python --use_swin --name swin_p0.5_circle_w5_b16_lr0.01_DG --lr 0.01 --batch 16 --DG --erasing_p 0.5 --circle --warm_epoch 5; python --name swin_p0.5_circle_w5_b16_lr0.01_DG
  • More training iterations may lead to better results.
  • Swin costs more GPU memory (11G GPU is needed) to run.
  • The hyper-parameter of DG-Market --DG is not tuned. Better hyper-parameter may lead to better results.

Different Losses

I do not optimize the hyper-parameters. You are free to tune them for better performance.

Methods Rank@1 mAP Reference
CE 92.01% 79.31% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_100 --total 100 ; python --name warm5_s1_b32_lr8_p0.5_100
CE + Sphere [Paper] 92.01% 79.39% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_sphere100 --sphere --total 100; python --name warm5_s1_b32_lr8_p0.5_sphere100
CE + Triplet [Paper] 92.40% 79.71% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_triplet100 --triplet --total 100; python --name warm5_s1_b32_lr8_p0.5_triplet100
CE + Lifted [Paper] 91.78% 79.77% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_lifted100 --lifted --total 100; python --name warm5_s1_b32_lr8_p0.5_lifted100
CE + Instance [Paper] 92.73% 81.11% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_instance100_gamma64 --instance --ins_gamma 64 --total 100 ; python --name warm5_s1_b32_lr8_p0.5_instance100_gamma64
CE + Contrast [Paper] 92.28% 81.42% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_contrast100 --contrast --total 100; python --name warm5_s1_b32_lr8_p0.5_contrast100
CE + Circle [Paper] 92.46% 81.70% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_circle100 --circle --total 100 ; python --name warm5_s1_b32_lr8_p0.5_circle100
CE + Contrast + Sphere 92.79% 82.02% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 32 --lr 0.08 --name warm5_s1_b32_lr8_p0.5_cs100 --contrast --sphere --total 100; python --name warm5_s1_b32_lr8_p0.5_cs100
CE + Contrast + Triplet (Long) 92.61% 82.01% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 24 --lr 0.062 --name warm5_s1_b24_lr6.2_p0.5_contrast_triplet_133 --contrast --triplet --total 133 ; python --name warm5_s1_b24_lr6.2_p0.5_contrast_triplet_133
CE + Contrast + Circle (Long) 92.19% 82.07% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 24 --lr 0.08 --name warm5_s1_b24_lr8_p0.5_contrast_circle133 --contrast --circle --total 133 ; python --name warm5_s1_b24_lr8_p0.5_contrast_circle133
CE + Contrast + Sphere (Long) 92.84% 82.37% python --warm_epoch 5 --stride 1 --erasing_p 0.5 --batchsize 24 --lr 0.06 --name warm5_s1_b24_lr6_p0.5_contrast_sphere133 --contrast --sphere --total 133 ; python --name warm5_s1_b24_lr6_p0.5_contrast_sphere133

Model Structure

You may learn more from We add one linear layer(bottleneck), one batchnorm layer and relu.


  • Python 3.6+
  • GPU Memory >= 6G
  • Numpy
  • Pytorch 0.3+
  • timm pip install timm for Swin-Transformer with Pytorch >1.7.0
  • pretrainedmodels via pip install pretrainedmodels
  • [Optional] apex (for float16)
  • [Optional] pretrainedmodels

(Some reports found that updating numpy can arrive the right accuracy. If you only get 50~80 Top1 Accuracy, just try it.) We have successfully run the code based on numpy 1.12.1 and 1.13.1 .

Getting started


  • Install Pytorch from
  • Install Torchvision from the source
    git clone
    cd vision
    python install
  • [Optional] You may skip it. Install apex from the source
    git clone
    cd apex
    python install --cuda_ext --cpp_ext

    Because pytorch and torchvision are ongoing projects.

Here we noted that our code is tested based on Pytorch 0.3.0/0.4.0/0.5.0/1.0.0 and Torchvision 0.2.0/0.2.1 .

Dataset & Preparation

Download Market1501 Dataset [Google] [Baidu] Or use command line:

pip install gdown 
pip install --upgrade gdown #!!important!!
gdown 0B8-rUzbwVRk0c054eEozWG9COHM

Preparation: Put the images with the same id in one folder. You may use


Remember to change the dataset path to your own path.

Futhermore, you also can test our code on DukeMTMC-reID Dataset( GoogleDriver or (BaiduYun password: bhbh)) Or use command line:

gdown 1jjE85dRCMOgRtvJ5RQV9-Afs-2_5dY3O

Our baseline code is not such high on DukeMTMC-reID Rank@1=64.23%, mAP=43.92%. Hyperparameters are need to be tuned.

  • [Optional] DG-Market is a generated pedestrian dataset of 128,307 images for training a robust model.


Train a model by

python --gpu_ids 0 --name ft_ResNet50 --train_all --batchsize 32  --data_dir your_data_path

--gpu_ids which gpu to run.

--name the name of model.

--data_dir the path of the training data.

--train_all using all images to train.

--batchsize batch size.

--erasing_p random erasing probability.

Train a model with random erasing by

python --gpu_ids 0 --name ft_ResNet50 --train_all --batchsize 32  --data_dir your_data_path --erasing_p 0.5


Use trained model to extract feature by

python --gpu_ids 0 --name ft_ResNet50 --test_dir your_data_path  --batchsize 32 --which_epoch 59

--gpu_ids which gpu to run.

--batchsize batch size.

--name the dir name of trained model.

--which_epoch select the i-th model.

--data_dir the path of the testing data.



It will output Rank@1, Rank@5, Rank@10 and mAP results. You may also try to conduct a faster evaluation with GPU.

For mAP calculation, you also can refer to the C++ code for Oxford Building. We use the triangle mAP calculation (consistent with the Market1501 original code).



It may take more than 10G Memory to run. So run it on a powerful machine if possible.

It will output Rank@1, Rank@5, Rank@10 and mAP results.


Notes the format of the camera id and the number of cameras.

For some dataset, e.g., MSMT17, there are more than 10 cameras. You need to modify the and to read the double-digit camera ID.

For some vehicle re-ID datasets. e.g. VeRi, you also need to modify the and It has different naming rules. (Sorry. It is in Chinese)


The following paper uses and reports the result of the baseline model. You may cite it in your paper.

  title={Joint discriminative and generative learning for person re-identification},
  author={Zheng, Zhedong and Yang, Xiaodong and Yu, Zhiding and Zheng, Liang and Yang, Yi and Kautz, Jan},
  journal={IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},

The following papers may be the first two to use the bottleneck baseline. You may cite them in your paper.

  author    = {Yifan Sun and
               Liang Zheng and
               Weijian Deng and
               Shengjin Wang},
  title     = {SVDNet for Pedestrian Retrieval},
  booktitle   = {ICCV},
  year      = {2017},

  title={In Defense of the Triplet Loss for Person Re-Identification},
  author={Hermans, Alexander and Beyer, Lucas and Leibe, Bastian},
  journal={arXiv preprint arXiv:1703.07737},

Basic Model

  title={A discriminatively learned CNN embedding for person reidentification},
  author={Zheng, Zhedong and Zheng, Liang and Yang, Yi},
  journal={ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM)},

  title={VehicleNet: Learning Robust Visual Representation for Vehicle Re-identification},
  author={Zheng, Zhedong and Ruan, Tao and Wei, Yunchao and Yang, Yi and Mei, Tao},
  journal={IEEE Transaction on Multimedia (TMM)},
  1. Pedestrian Alignment Network GitHub stars
  2. 2stream Person re-ID GitHub stars
  3. Pedestrian GAN GitHub stars
  4. Language Person Search GitHub stars
  5. DG-Net GitHub stars
  6. 3D Person re-ID GitHub stars