Home

Awesome

IJCV 2024: Transformer-based ReID Survey

Transformer for Object Re-Identification: A Survey. arXiv

Highlights

Citation

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

@article{ye2024transformer,
  title={Transformer for Object Re-Identification: A Survey},
  author={Ye, Mang and Chen, Shuoyi and Li, Chenyue and Zheng, Wei-Shi and Crandall, David and Du, Bo},
  journal={arXiv preprint arXiv:2401.06960},
  year={2024}
}

TPAMI 2021 ReID-Survey with a Powerful AGW Baseline

Deep Learning for Person Re-identification: A Survey and Outlook. PDF with supplementary materials. arXiv

A simplified introduction in Chinese on 知乎.

Highlights

AGW on Single-Modality Image Re-ID with mINP

DukeMTMC dataset

MethodPretrainedRank@1mAPmINPModelPaper
BagTricksImageNet86.4%76.4%40.7%CodeBag of Tricks and A Strong Baseline for Deep Person Re-identification. In ArXiv 19. PDF
ABD-NetImageNet89.0%78.6%42.1%CodeABD-Net: Attentive but Diverse Person Re-Identification. In ICCV 19. PDF
AGWImageNet89.0%79.6%45.7%GoogleDriveDeep Learning for Person Re-identification: A Survey and Outlook

Market-1501 dataset

MethodPretrainedRank@1mAPmINPModelPaper
BagTricksImageNet94.5%85.9%59.4%CodeBag of Tricks and A Strong Baseline for Deep Person Re-identification. In ArXiv 19. arXiv
ABD-NetImageNet95.6%88.3%66.2%CodeABD-Net: Attentive but Diverse Person Re-Identification. In ICCV 19. PDF
AGWImageNet95.1%87.8%65.0%GoogleDriveDeep Learning for Person Re-identification: A Survey and Outlook. In ArXiv 20. arXiv

CUHK03 dataset

MethodPretrainedRank@1mAPmINPModelPaper
BagTricksImageNet58.0%56.6%43.8%CodeBag of Tricks and A Strong Baseline for Deep Person Re-identification. In ArXiv 19. PDF
AGWImageNet63.6%62.0%50.3%GoogleDriveDeep Learning for Person Re-identification: A Survey and Outlook. In ArXiv 20. arXiv

MSMT17 dataset

MethodPretrainedRank@1mAPmINPModelPaper
BagTricksImageNet63.4%45.1%12.4%CodeBag of Tricks and A Strong Baseline for Deep Person Re-identification. In ArXiv 19. arXiv
AGWImageNet68.3%49.3%14.7%GoogleDriveDeep Learning for Person Re-identification: A Survey and Outlook. In ArXiv 20. arXiv

Quick Start

1. Prepare dataset

Create a directory to store reid datasets under this repo, taking Market1501 for example

cd ReID-Survey
mkdir toDataset
toDataset
    market1501 
        bounding_box_test/
        bounding_box_train/
        ......

Partial-REID and Partial-iLIDS datasets are provided by https://github.com/lingxiao-he/Partial-Person-ReID

2. Install dependencies

3. Train

To train a AGW model with on Market1501 with GPU device 0, run similarly:

python3 tools/main.py --config_file='configs/AGW_baseline.yml' MODEL.DEVICE_ID "('0')" DATASETS.NAMES "('market1501')" OUTPUT_DIR "('./log/market1501/Experiment-AGW-baseline')"

4. Test

To test a AGW model with on Market1501 with weight file './pretrained/dukemtmc_AGW.pth', run similarly:

python3 tools/main.py --config_file='configs/AGW_baseline.yml' MODEL.DEVICE_ID "('0')" DATASETS.NAMES "('market1501')"  MODEL.PRETRAIN_CHOICE "('self')" TEST.WEIGHT "('./pretrained/market1501_AGW.pth')" TEST.EVALUATE_ONLY "('on')" OUTPUT_DIR "('./log/Test')"

Citation

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

@article{pami21reidsurvey,
  title={Deep Learning for Person Re-identification: A Survey and Outlook},
  author={Ye, Mang and Shen, Jianbing and Lin, Gaojie and Xiang, Tao and Shao, Ling and Hoi, Steven C. H.},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year={2021},
}

Contact: mangye16@gmail.com