Nothing Special   »   [go: up one dir, main page]
Skip to main content
Springer Nature Link
Log in

Dual semantic-aligned clustering for cross-domain person re-identification

  • Regular Paper
  • Published:
Multimedia Systems Aims and scope Submit manuscript

Abstract

Person re-identification (ReID) intends to link people with common identities captured by various cameras spread across a big, fragmented area. Due to the imbalance in the distribution of data between input and output variables and the absence of labeled information in the target domain, cross-domain person re-identification (ReID) is a high demanding assignment. This paper presents a contemporary dual semantic-aligned (DSA) clustering approach, which jointly learns latent semantic subspaces of source and target domains to group target samples. More specifically, we present an innovative subspace learning model which jointly finds intra-domain basis and inter-domain multi-granularity basis. After the subspaces are learned, target samples are correspondingly projected and clustered automatically in each subspace. The clusters allow the model to attribute imaginary labels to the target samples and then impose supervised constraints throughout the training phase. Experiment results on DukeMTMC-reID, Market-1501, and MSMT17 benchmarks validate that our scheme significantly outperformed the state-of-the-art schemes. Our approach obtains performance of 86.82%/63.06% (Duke\(\rightarrow\)Market), 76.3%/58.35% (Market\(\rightarrow\)Duke) and 52.97%/22.44% (Duke\(\rightarrow\)MSMT17) in Rank-1/mAP, correspondingly.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

References

  1. Aharon, M., Elad, M., Bruckstein, A.: K-svd: an algorithm for designing overcomplete dictionaries for sparse representation. IEEE Trans Signal Proc 54(11), 4311–4322 (2006)

    Article  MATH  Google Scholar 

  2. Bai Z, Wang Z, Wang J, et al (2021) Unsupervised multi-source domain adaptation for person re-identification. In: CVPR

  3. Cao Z, Simon T, Wei SE, et al (2017) Realtime multi-person 2d pose estimation using part affinity fields. In: CVPR

  4. Chang X, Yang Y, Xiang T, et al (2019) Disjoint label space transfer learning with common factorised space. In: AAAI

  5. Chen H, Lagadec B, Bremond F (2021a) Ice: inter-instance contrastive encoding for unsupervised person re-identification. In: ICCV

  6. Chen H, Wang Y, Lagadec B, et al (2021b) Joint generative and contrastive learning for unsupervised person re-identification. In: CVPR

  7. Chen Y, Zhu X, Gong S (2019) Instance-guided context rendering for cross-domain person re-identification. In: ICCV

  8. Cho Y, Kim WJ, Hong S, et al (2022) Part-based pseudo label refinement for unsupervised person re-identification. In: CVPR

  9. Deng J, Dong W, Socher R, et al (2009) Imagenet: a large-scale hierarchical image database. In: CVPR

  10. Deng W, Zheng L, Ye Q, et al (2018) Image-image domain adaptation with preserved self-similarity and domain-dissimilarity for person reidentification. In: CVPR

  11. Ester M, Kriegel HP, Sander J, et al (1996) A density-based algorithm for discovering clusters a density-based algorithm for discovering clusters in large spatial databases with noise. In: KDD

  12. Fan, H., Zheng, L., Yang, Y.: Unsupervised person re-identification: clustering and fine-tuning. ACM Trans Multimed Comput Commun Appl 14(4), 1–8 (2018)

    Article  Google Scholar 

  13. Fu Y, Wei Y, Wang G, et al (2019) Self-similarity grouping: a simple unsupervised cross domain adaptation approach for person re-identification. In: ICCV

  14. Goodfellow IJ, Pouget-Abadie J, Mirza M, et al (2014) Generative adversarial nets. In: NeurIPS

  15. Guler RA, Neverova N, Kokkinos I (2018) Densepose: dense human pose estimation in the wild. In: CVPR

  16. He K, Zhang X, Ren S, et al (2016) Deep residual learning for image recognition. In: CVPR

  17. Hermans A, Beyer L, Leibe B (2017) In defense of the triplet loss for person re-identification. arXiv preprint arXiv:1703.07737

  18. Huang H, Yang W, Chen X, et al (2018) Eanet: enhancing alignment for cross-domain person re-identification. arXiv preprint arXiv:1812.11369

  19. Huang Y, Wu Q, Xu J, et al (2019) Sbsgan: suppression of inter-domain background shift for person. In: ICCV

  20. Isobe T, Li D, Tian L, et al (2021) Towards discriminative representation learning for unsupervised person re-identification. In: ICCV

  21. Ji H, Wang L, Zhou S, et al (2021) Meta pairwise relationship distillation for unsupervised person re-identification. In: ICCV

  22. Kalayeh MM, Basaran E, Gokmen M, et al (2018) Human semantic parsing for person re-identification. In: CVPR

  23. Kodirov E, Xiang T, Gong S (2015) Dictionary learning with iterative laplacian regularisation for unsupervised person reidentification. In: BMVC

  24. Li, H., Dong, N., Yu, Z., et al.: Triple adversarial learning and multi-view imaginative reasoning for unsupervised domain adaptation person re-identification. IEEE Trans Circuits Syst Video Technol 32(5), 2814–2830 (2022)

    Article  MathSciNet  Google Scholar 

  25. Li W, Zhu X, Gong S (2018) Harmonious attention network for person re-identification. In: CVPR

  26. Li YJ, Lin CS, Lin YB, et al (2019) Cross-dataset person re-identification via unsupervised pose disentanglement and adaptation. In: ICCV

  27. Liao S, Hu Y, Zhu X, et al (2015) Person re-identification by local maximal occurrence representation and metric learning. In: CVPR

  28. Lin S, Li H, Li CT, et al (2018) Multi-task mid-level feature alignment network for unsupervised cross-dataset person re-identification. In: BMVC

  29. Lin Y, Xie L, Wu Y, et al (2020) Unsupervised person re-identification via softened similarity learning. In: CVPR

  30. Lisanti, G., Masi, I., Bagdanov, A.D., et al.: Person re-identification by iterative re-weighted sparse ranking. IEEE Trans Patt Analy Mach Intell 37(8), 1629–1642 (2015)

    Article  Google Scholar 

  31. Liu J, Zha ZJ, Chen D, et al (2019) Adaptive transfer network for cross-domain person re-identification. In: CVPR

  32. Long M, Cao Y, Wang J, et al (2015) Learning transferable features with deep adaptation networks. In: ICML

  33. Lv J, Chen W, Li Q, et al (2018) Unsupervised cross-dataset person re-identification by transfer learning of spatial-temporal patterns. In: CVPR

  34. Qi L, Wang L, Huo J, et al (2019) A novel unsupervised camera-aware domain adaptation framework for person re-identification. In: ICCV

  35. Qian X, Fu Y, Xiang T, et al (2018) Pose-normalized image generation for person re-identification. In: ECCV

  36. Rubinstein, R., Bruckstein, M.A., et al.: Dictionaries for sparse representation modeling. Proc IEEE 98(6), 1045–1057 (2010)

    Article  Google Scholar 

  37. Song L, Wang C, Zhang L, et al (2018) Unsupervised domain adaptive re-identification: theory and practice. arXiv preprint arXiv:1807.11334

  38. Su C, Zhang S, Xing J, et al (2016) Deep attributes driven multi-camera person re-identification. In: ECCV

  39. Sun B, Saenko K (2016) Deep coral: correlation alignment for deep domain adaptation. In: ECCV

  40. Torralba A, Efros AA (2011) Unbiased look at dataset bias. In: CVPR

  41. Verma A, Subramanyam AV, Wang Z, et al (2021) Unsupervised domain adaptation for person re-identification via individual-preserving and environmental-switching cyclic generation. IEEE Transactions on Multimedia. vol. 25, pp. 364–377

  42. Wang D, Zhang S (2020) Unsupervised person re-identification via multi-label classification. In: CVPR

  43. Wang G, Lai JH, Liang W, et al (2020) Smoothing adversarial domain attack and p-memory reconsolidation for cross-domain person re-identification. In: CVPR

  44. Wang J, Zhu X, Gong S, et al (2018) Transferable joint attribute-identity deep learning for unsupervised person re-identification. In: CVPR

  45. Wei L, Zhang S, Gao W, et al (2018) Person transfer gan to bridge domain gap for person re-identification. In: CVPR

  46. Xuan S, Zhang S (2021) Intra-inter camera similarity for unsupervised person re-identification. In: CVPR

  47. Yang, F., Yan, K., Lu, S., et al.: Attention driven person re-identification. Patt Recogn 86, 143–155 (2019)

    Article  Google Scholar 

  48. Yang F, Zhong Z, Luo Z, et al (2021a) Joint noise-tolerant learning and meta camera shift adaptation for unsupervised person re-identification. In: CVPR

  49. Yang F, Zhong Z, Luo Z, et al (2021b) Joint noise-tolerant learning and meta camera shift adaptation for unsupervised person re-identification. In: CVPR

  50. Yang Q, Yu HX, Wu A, et al (2019b) Patch-based discriminative feature learning for unsupervised person re-identification. In: CVPR

  51. Yu HX, Wu A, Zheng WS (2017) Cross-view asymmetric metric learning for unsupervised person re-identification. In: ICCV

  52. Yu HX, Zheng WS, Wu A, et al (2019) Unsupervised person re-identification by soft multilabel learning. In: CVPR

  53. Zeng K, Ning M, Wang Y, et al (2020) Hierarchical clustering with hard-batch triplet loss for person re-identification. In: CVPR

  54. Zhang, H., Cao, H., Yang, X., et al.: Self-training with progressive representation enhancement for unsupervised cross-domain person re-identification. IEEE Trans Image Proc 30, 5287–5298 (2021)

    Article  Google Scholar 

  55. Zhang X, Cao J, Shen C, et al (2019) Self-training with progressive augmentation for unsupervised cross-domain person re-identification. In: ICCV

  56. Zhang X, Li D, Wang Z, et al (2022) Implicit sample extension for unsupervised person re-identification. In: CVPR

  57. Zheng L, Shen L, Tian L, et al (2015) Scalable person re-identification: a benchmark. In: ICCV

  58. Zheng Z, Zheng L, Yang Y (2017) Unlabeled samples generated by gan improve the person re-identification baseline in vitro. In: ICCV

  59. Zhong Z, Zheng L, Li S, et al (2018) Generalizing a person retrieval model hetero- and homogeneously. In: ECCV

  60. Zhong Z, Zheng L, Luo Z, et al (2019a) Invariance matters: exemplar memory for domain adaptive person re-identification. In: CVPR

  61. Zhong, Z., Zheng, L., Zheng, Z., et al.: Camstyle: a novel data augmentation method for person re-identification. IEEE Trans Image Proc 28(3), 1176–1190 (2019)

    Article  MathSciNet  Google Scholar 

  62. Zhong, Z., Zheng, L., Luo, Z., et al.: Learning to adapt invariance in memory for person re-identification. IEEE Trans Patt Analy Mach Intell 43(8), 2723–2738 (2021)

    Google Scholar 

Download references

Acknowledgements

This work was supported in part by the Macao Science and Technology Development Fund Project under contract No.0005/2021/AIR.

Author information

Authors and Affiliations

  1. Macau University of Science and Technology, Avenida Wai Long, Macau, 999078, China

    Changshui Yang, Yan Liu, Qiang Liu & Wenyong Wang

  2. Yangtze Delta Region Institute (Huzhou), University of Electronic Science and Technology of China, Xisaishan Road, Huzhou, 313001, Zhejiang, China

    Riaz Ullah Khan

  3. Fnetlink Tecnology Company Limited, Della Sinagna Street, Macau, 999078, China

    Bin Chen

Authors
  1. Changshui Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yan Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Qiang Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Riaz Ullah Khan
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Bin Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Wenyong Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

All the authors discussed the idea of the this work, as well as revised and reviewed the manuscript.

Corresponding author

Correspondence to Wenyong Wang.

Ethics declarations

Conflict of interest

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yang, C., Liu, Y., Liu, Q. et al. Dual semantic-aligned clustering for cross-domain person re-identification. Multimedia Systems 29, 2351–2362 (2023). https://doi.org/10.1007/s00530-023-01111-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00530-023-01111-z

Keywords

Search

Navigation