Global Distance-Distributions Separation for Unsupervised Person Re-identification

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 12352)


Supervised person re-identification (ReID) often has poor scalability and usability in real-world deployments due to domain gaps and the lack of annotations for the target domain data. Unsupervised person ReID through domain adaptation is attractive yet challenging. Existing unsupervised ReID approaches often fail in correctly identifying the positive samples and negative samples through the distance-based matching/ranking. The two distributions of distances for positive sample pairs (Pos-distr) and negative sample pairs (Neg-distr) are often not well separated, having large overlap. To address this problem, we introduce a global distance-distributions separation (GDS) constraint over the two distributions to encourage the clear separation of positive and negative samples from a global view. We model the two global distance distributions as Gaussian distributions and push apart the two distributions while encouraging their sharpness in the unsupervised training process. Particularly, to model the distributions from a global view and facilitate the timely updating of the distributions and the GDS related losses, we leverage a momentum update mechanism for building and maintaining the distribution parameters (mean and variance) and calculate the loss on the fly during the training. Distribution-based hard mining is proposed to further promote the separation of the two distributions. We validate the effectiveness of the GDS constraint in unsupervised ReID networks. Extensive experiments on multiple ReID benchmark datasets show our method leads to significant improvement over the baselines and achieves the state-of-the-art performance.


Unsupervised learning Person re-identification Global distance-distributions separation Momentum update Hard mining 



This work was supported in part by NSFC under Grant U1908209, 61632001 and the National Key Research and Development Program of China 2018AAA0101400.

Supplementary material

504444_1_En_43_MOESM1_ESM.pdf (614 kb)
Supplementary material 1 (pdf 614 KB)


  1. 1.
    Ahmed, E., Jones, M., Marks, T.K.: An improved deep learning architecture for person re-identification. In: CVPR (2015)Google Scholar
  2. 2.
    Almazan, J., Gajic, B., Murray, N., Larlus, D.: Re-id done right: towards good practices for person re-identification. arXiv preprint arXiv:1801.05339 (2018)
  3. 3.
    Chen, L., Ai, H., Zhuang, Z., Shang, C.: Real-time multiple people tracking with deeply learned candidate selection and person re-identification. In: ICME, pp. 1–6 (2018)Google Scholar
  4. 4.
    Davis, J., Goadrich, M.: The relationship between precision-recall and roc curves. In: ICML, pp. 233–240 (2006)Google Scholar
  5. 5.
    Deng, W., Zheng, L., Ye, Q., Kang, G., Yang, Y., Jiao, J.: Image-image domain adaptation with preserved self-similarity and domain-dissimilarity for person re-identification. In: CVPR (2018)Google Scholar
  6. 6.
    Ester, M., Kriegel, H.P., Sander, J., Xu, X., et al.: A density-based algorithm for discovering clusters in large spatial databases with noise. Kdd. 96, 226–231 (1996)Google Scholar
  7. 7.
    Fan, H., Zheng, L., Yan, C., Yang, Y.: Unsupervised person re-identification: clustering and fine-tuning. ACM Trans. Multimedia Comput. Commun. Appl. (TOMM) 14(4), 1–18 (2018)CrossRefGoogle Scholar
  8. 8.
    Fawcett, T.: An introduction to ROC analysis. Pattern Recogn. Lett. 27(8), 861–874 (2006)MathSciNetCrossRefGoogle Scholar
  9. 9.
    Fu, Y., Wei, Y., Wang, G., Zhou, X., Shi, H., Huang, T.S.: Self-similarity grouping: a simple unsupervised cross domain adaptation approach for person re-identification. In: ICCV (2019)Google Scholar
  10. 10.
    Fu, Y., Wei, Y., Zhou, Y., et al.: Horizontal pyramid matching for person re-identification. In: AAAI (2019)Google Scholar
  11. 11.
    Ge, Y., Li, Z., Zhao, H., et al.: FD-GAN: pose-guided feature distilling gan for robust person re-identification. In: NeurIPS (2018)Google Scholar
  12. 12.
    Grafarend, E., Awange, J.: Linear and Nonlinear Models. Springer, Heidelberg (2012). Scholar
  13. 13.
    Hermans, A., Beyer, L., Leibe, B.: In defense of the triplet loss for person re-identification. arXiv preprint arXiv:1703.07737 (2017)
  14. 14.
    Hirzer, M., Roth, P.M., Köstinger, M., Bischof, H.: Relaxed pairwise learned metric for person re-identification. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012. LNCS, vol. 7577, pp. 780–793. Springer, Heidelberg (2012). Scholar
  15. 15.
    Jia, M., Zhai, Y., Lu, S., Ma, S., Zhang, J.: A similarity inference metric for RGB-infrared cross-modality person re-identification. In: IJCAI (2020)Google Scholar
  16. 16.
    Jin, X., Lan, C., Zeng, W., Chen, Z.: Uncertainty-aware multi-shot knowledge distillation for image-based object re-identification. In: AAAI (2020)Google Scholar
  17. 17.
    Jin, X., Lan, C., Zeng, W., Chen, Z., Zhang, L.: Style normalization and restitution for generalizable person re-identification. In: CVPR (2020)Google Scholar
  18. 18.
    Jin, X., Lan, C., Zeng, W., Wei, G., Chen, Z.: Semantics-aligned representation learning for person re-identification. In: AAAI, pp. 11173–11180 (2020)Google Scholar
  19. 19.
    Kumar, B., Carneiro, G., Reid, I., et al.: Learning local image descriptors with deep siamese and triplet convolutional networks by minimising global loss functions. In: CVPR, pp. 5385–5394 (2016)Google Scholar
  20. 20.
    Li, M., Zhu, X., Gong, S.: Unsupervised person re-identification by deep learning tracklet association. In: ECCV (2018)Google Scholar
  21. 21.
    Li, W., Zhao, R., Tian, L., et al.: Deepreid: deep filter pairing neural network for person re-identification. In: CVPR (2014)Google Scholar
  22. 22.
    Lin, S., Li, H., Li, C.T., Kot, A.C.: Multi-task mid-level feature alignment network for unsupervised cross-dataset person re-identification. In: BMVC (2018)Google Scholar
  23. 23.
    D Liu, J., Zha, Z.J., Chen, D., Hong, R., Wang, M.: Adaptive transfer network for cross-domain person re-identification. In: CVPR (2019)Google Scholar
  24. 24.
    Luo, H., Gu, Y., Liao, X., Lai, S., Jiang, W.: Bag of tricks and a strong baseline for deep person re-identification. In: CVPR workshops (2019)Google Scholar
  25. 25.
    Qi, L., Wang, L., Huo, J., Zhou, L., Shi, Y., Gao, Y.: A novel unsupervised camera-aware domain adaptation framework for person re-identification. In: ICCV (2019)Google Scholar
  26. 26.
    Qian, X., Fu, Y., Wang, W., et al.: Pose-normalized image generation for person re-identification. In: ECCV (2018)Google Scholar
  27. 27.
    Ridgeway, K., Mozer, M.C.: Learning deep disentangled embeddings with the f-statistic loss. In: NeurIPS, pp. 185–194 (2018)Google Scholar
  28. 28.
    Ristani, E., Tomasi, C.: Features for multi-target multi-camera tracking and re-identification. In: CVPR, pp. 6036–6046 (2018)Google Scholar
  29. 29.
    Song, L., et al.: Unsupervised domain adaptive re-identification: theory and practice. Pattern Recogn. 102, 107173 (2020)CrossRefGoogle Scholar
  30. 30.
    Su, C., Li, J., Zhang, S., et al.: Pose-driven deep convolutional model for person re-identification. In: ICCV (2017)Google Scholar
  31. 31.
    Sun, Y., Zheng, L., Yang, Y., Tian, Q., Wang, S.: Beyond part models: person retrieval with refined part pooling (and a strong convolutional baseline). In: ECCV, pp. 480–496 (2018)Google Scholar
  32. 32.
    Tang, H., Zhao, Y., Lu, H.: Unsupervised person re-identification with iterative self-supervised domain adaptation. In: CVPR workshops (2019)Google Scholar
  33. 33.
    Ustinova, E., Lempitsky, V.: Learning deep embeddings with histogram loss. In: NeurIPS, pp. 4170–4178 (2016)Google Scholar
  34. 34.
    Wang, J., Zhu, X., Gong, S., Li, W.: Transferable joint attribute-identity deep learning for unsupervised person re-identification. In: CVPR (2018)Google Scholar
  35. 35.
    Wei, L., Zhang, S., Gao, W., Tian, Q.: Person transfer GAN to bridge domain gap for person re-identification. In: CVPR (2018)Google Scholar
  36. 36.
    Wojke, N., Bewley, A.: Deep cosine metric learning for person re-identification. In: WACV, pp. 748–756 (2018)Google Scholar
  37. 37.
    Yang, F., et al.: Asymmetric co-teaching for unsupervised cross domain person re-identification. In: AAAI (2020)Google Scholar
  38. 38.
    Yang, Q., Yu, H.X., Wu, A., Zheng, W.S.: Patch-based discriminative feature learning for unsupervised person re-identification. In: CVPR (2019)Google Scholar
  39. 39.
    Ye, M., Shen, J., Lin, G., Xiang, T., Shao, L., Hoi, S.C.: Deep learning for person re-identification: a survey and outlook. arXiv preprint arXiv:2001.04193 (2020)
  40. 40.
    Yu, H.X., Wu, A., Zheng, W.S.: Cross-view asymmetric metric learning for unsupervised person re-identification. In: ICCV (2017)Google Scholar
  41. 41.
    Yu, H.X., Wu, A., Zheng, W.S.: Unsupervised person re-identification by deep asymmetric metric embedding. In: IEEE TPAMI (2018)Google Scholar
  42. 42.
    Yu, H.X., Zheng, W.S., Wu, A., Guo, X., Gong, S., Lai, J.H.: Unsupervised person re-identification by soft multilabel learning. In: CVPR (2019)Google Scholar
  43. 43.
    Zhai, Y., et al.: Ad-cluster: augmented discriminative clustering for domain adaptive person re-identification. In: CVPR, pp. 9021–9030 (2020)Google Scholar
  44. 44.
    Zhai, Y., Ye, Q., Lu, S., Jia, M., Ji, R., Tian, Y.: Multiple expert brainstorming for domain adaptive person re-identification. In: ECCV (2020)Google Scholar
  45. 45.
    Zhang, X., Cao, J., Shen, C., You, M.: Self-training with progressive augmentation for unsupervised cross-domain person re-identification. In: ICCV (2019)Google Scholar
  46. 46.
    Zhang, Z., Lan, C., Zeng, W., et al.: Densely semantically aligned person re-identification. In: CVPR (2019)Google Scholar
  47. 47.
    Zhao, H., Tian, M., Sun, S., et al.: Spindle net: person re-identification with human body region guided feature decomposition and fusion. In: CVPR (2017)Google Scholar
  48. 48.
    Zheng, L., Shen, L., et al.: Scalable person re-identification: a benchmark. In: ICCV (2015)Google Scholar
  49. 49.
    Zheng, L., Yang, Y., Hauptmann, A.G.: Person re-identification: past, present and future. arXiv preprint arXiv:1610.02984 (2016)
  50. 50.
    Zheng, L., Zhang, H., Sun, S., Chandraker, M., Yang, Y., Tian, Q.: Person re-identification in the wild. In: CVPR, pp. 1367–1376 (2017)Google Scholar
  51. 51.
    Zheng, Z., Zheng, L., Yang, Y.: Unlabeled samples generated by gan improve the person re-identification baseline in vitro. In: ICCV (2017)Google Scholar
  52. 52.
    Zhong, Z., Zheng, L., Cao, D., Li, S.: Re-ranking person re-identification with k-reciprocal encoding. In: CVPR (2017)Google Scholar
  53. 53.
    Zhong, Z., Zheng, L., Li, S., Yang, Y.: Generalizing a person retrieval model hetero-and homogeneously. In: ECCV (2018)Google Scholar
  54. 54.
    Zhong, Z., Zheng, L., Luo, Z., Li, S., Yang, Y.: Invariance matters: exemplar memory for domain adaptive person re-identification. In: CVPR, pp. 598–607 (2019)Google Scholar
  55. 55.
    Zhou, S., Wang, J., Hou, Q., Gong, Y.: Deep ranking model for person re-identification with pairwise similarity comparison. In: Chen, E., Gong, Y., Tie, Y. (eds.) PCM 2016. LNCS, vol. 9917, pp. 84–94. Springer, Cham (2016). Scholar
  56. 56.
    Zhu, J.Y., Park, T., Isola, P., et al.: Unpaired image-to-image translation using cycle-consistent adversarial networks. In: ICCV (2017)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.University of Science and Technology of ChinaHefeiChina
  2. 2.Microsoft Research AsiaBeijingChina

Personalised recommendations