Awecome-Domain-Generalizable-Person-Re-identification

July 1, 2025 · View on GitHub

"Domain Generalization for Person Re-identification: A Survey Towards Domain-Agnostic Person Matching" (Neurocomputing 2025).

awesome Visitors arXiv Stars

📚 Contents

📄 Paper List

We categorize DG-ReID papers based on their methodologies:

Normalization-based

Title Publication Date
Style Normalization and Restitution for Generalizable Person Re-IdentificationCVPR2020
Meta Batch-Instance Normalization for Generalizable Person Re-IdentificationCVPR2021
Adaptive Cross-Domain Learning for Generalizable Person Re-IdentificationECCV2022
Dynamically Transformed Instance Normalization Network for Generalizable Person Re-IdentificationECCV2022
Rethinking Normalization Layers for Domain Generalizable Person Re-identificationECCV2022

Mixture-of-Experts-based

Title Publication Date
Generalizable Person Re-identification with Relevance-aware Mixture of ExpertsCVPR2021
Mimic Embedding via Adaptive Aggregation: Learning Generalizable Person Re-identificationECCV2022

Memory-based

Title Publication Date
Generalizable Person Re-identification by Domain-Invariant Mapping NetworkCVPR2019
Interpretable and Generalizable Person Re-identification with Query-adaptive Convolution and Temporal LiftingECCV2020
Part-Aware Transformer for Generalizable Person Re-identificationICCV2023

Meta-learning-based

Title Publication Date
Person30K: A Dual-Meta Generalization Network for Person Re-IdentificationCVPR2021
Learning to Generalize Unseen Domains via Memory-based Multi-Source Meta-Learning for Person Re-IdentificationCVPR2021
Meta distribution alignment for generalizable person re-identificationCVPR2022

Data-driven learning-based

Title Publication Date
Cloning Outfits From Real-World Images to 3D Characters for Generalizable Person Re-IdentificationCVPR2022
Identity-Seeking Self-Supervised Representation Learning for Generalizable Person Re-identificationICCV2023
Generalizable Person Re-identification via Balancing Alignment and UniformityNeurIPS2024

CLIP-based

Title Publication Date
CLIP-DFGS: A Hard Sample Mining Method for CLIP in Generalizable Person Re-IdentificationACM TOMM2024
CILP-FGDI: Exploiting Vision-Language Model for Generalizable Person Re-IdentificationIEEE Transactions on Information Forensics and Security2025

Others

Title Publication Date
Mitigate Domain Shift by Primary-Auxiliary Objectives Association for Generalizing Person ReIDWACV2024
Diversity-Authenticity Co-constrained Stylization for Federated Domain Generalization in Person Re-identificationAAAI2024

📊 Datasets & Benchmarks

We include commonly used datasets for evaluating DG-ReID performance.

DatasetVenueType#ID#Image#CamURLPaper
CUHK01ACCV'12🔹 Train9713,8842🔗🔗 Human re-identification with transferred metric learning
CUHK02ACCV'12🔹 Train1,8167,26410🔗🔗 Human re-identification with transferred metric learning
CUHK03CVPR'14🔹 Train1,47613,1642🔗🔗 Deep filter pairing neural network for person re-identification
Market-1501ICCV'15🔹 Train1,50132,6886🔗🔗 Scalable person re-identification: A benchmark
CUHK-SYSUCVPR'17🔹 Train8,43218,1841🔗🔗 Joint detection and identification feature learning for person search
Duke-MTMCICCV'17🔹 Train1,40436,4118🔗🔗 Unlabeled samples generated by gan improve the person re-identification baseline in vitro
MSMT17CVPR'18🔹 Train4,101126,44115🔗🔗 Person transfer gan to bridge domain gap for person re-identification
PRID2011SCIA'11🔸 Test2001,1342🔗🔗 Person re-identification by descriptive and discriminative classification
VIPeRPETS'07🔸 Test6321,2642🔗🔗 Evaluating appearance models for recognition, reacquisition, and tracking
iLIDSECCV'18🔸 Test1194762🔗🔗 Unsupervised person re-identification by deeplearning tracklet association
GRIDICIP'13🔸 Test2501,2758🔗🔗 Person re-identification by manifold ranking

📣 News

  • 📌 2025-06: Paper accepted on Neurocomputing and repository initialized
  • 🚀 2025-05: Repository initialized.

🔍 Survey Paper

You can find the preprint of our survey here

The overview of our survey paper: taxonomy

🔖 Citation

If you find this survey helpful, please consider citing us:

@article{lee2025domain,
  title={Domain generalization for person re-identification: A survey towards domain-agnostic person matching},
  author={Lee, Hyeonseo and Park, Juhyun and Oh, Jihyong and Eom, Chanho},
  journal={Neurocomputing},
  pages={130763},
  year={2025},
  publisher={Elsevier}
}

💫 Star History

Star History Chart