Abstract
Person re-identification (Re-Id) is a beneficial computer vision functionality in security-related applications based on video surveillance systems. It is a challenging cross-camera matching problem, which makes it prone to domain shift issues. To mitigate them, supervised and unsupervised domain adaptation and domain generalisation (DG) methods have been proposed. All such methods tend to favour performance improvements on the target data set at the expense of performance on the source data set(s), on which they generally deteriorate. In this work, instead, we propose an alternative method for DG that does not involve any re-training or fine-tuning of the Re-Id model and thus has no adverse effect on the performance of the source data set. It exploits Generative Adversarial Networks, trained on the source data set only with a one-vs-all mapping that simulates the target data set images, with the aim of transferring the style of the source data set into the target images. Finally, an ad hoc ranking process combines the features extracted from the original and generated images and produces the final ranked list. The proposed method can be used on top of any Re-Id model, making it a possible alternative method against domain shift and also complementary to other approaches. The considered solution is evaluated on a challenging cross-data set scenario on two benchmark data sets and a deep learning baseline for Re-Id. The obtained results demonstrate that the proposed solution improves performance, especially when the Re-Id model is specialised in the source domain.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Ainam, J., et al.: Unsupervised domain adaptation for person re-identification with iterative soft clustering. Knowl.-Based Syst. 212, 106644 (2021). https://doi.org/10.1016/j.knosys.2020.106644
Chen, Y., Zhu, X., Gong, S.: Instance-guided context rendering for cross-domain person re-identification. In: ICCV, pp. 232–242 (2019). https://doi.org/10.1109/ICCV.2019.00032
Chong, Y., et al.: Style transfer for unsupervised domain-adaptive person re-identification. Neurocomputing 422, 314–321 (2021). https://doi.org/10.1016/j.neucom.2020.10.005
Delussu, R., Putzu, L., Fumera, G.: Human-in-the-loop cross-domain person re-identification. Expert Syst. Appl. 226, 120216 (2023). https://doi.org/10.1016/j.eswa.2023.120216
Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: Imagenet: a large-scale hierarchical image database. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 248–255 (2009)
Ding, G., Zhang, S., Khan, S.H., Tang, Z., Zhang, J., Porikli, F.: Feature affinity-based pseudo labeling for semi-supervised person re-identification. IEEE Trans. Multim. 21(11), 2891–2902 (2019). https://doi.org/10.1109/TMM.2019.2916456
Farenzena, M., et al.: Person re-identification by symmetry-driven accumulation of local features. In: CVPR, pp. 2360–2367 (2010). https://doi.org/10.1109/CVPR.2010.5539926
Genç, A., Ekenel, H.K.: Cross-dataset person re-identification using deep convolutional neural networks: effects of context and domain adaptation. Multimedia Tools Appl. 78(5), 5843–5861 (2018). https://doi.org/10.1007/s11042-018-6409-3
Gou, M., et al.: Dukemtmc4reid: a large-scale multi-camera person re-identification dataset. In: CVPR Workshops, pp. 1425–1434 (2017). https://doi.org/10.1109/CVPRW.2017.185
Han, H., Ma, W., Zhou, M., Guo, Q., Abusorrah, A.: A novel semi-supervised learning approach to pedestrian re-identification. IEEE Internet Things J. 8(4), 3042–3052 (2021). https://doi.org/10.1109/JIOT.2020.3024287
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Khatun, A., Denman, S., Sridharan, S., Fookes, C.: Pose-driven attention-guided image generation for person re-identification. Pattern Recogn. 137, 109246 (2023). https://doi.org/10.1016/j.patcog.2022.109246
Khosla, A., Zhou, T., Malisiewicz, T., Efros, A.A., Torralba, A.: Undoing the damage of dataset bias. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012. LNCS, vol. 7572, pp. 158–171. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-33718-5_12
Leng, Q., et al.: A survey of open-world person re-identification. IEEE T-CSVT 30(4), 1092–1108 (2020). https://doi.org/10.1109/TCSVT.2019.2898940
Lin, S., et al.: Multi-domain adversarial feature generalization for person re-identification. IEEE T-IP 30, 1596–1607 (2021). https://doi.org/10.1109/TIP.2020.3046864
Qi, L., Liu, J., Wang, L., Shi, Y., Geng, X.: Unsupervised generalizable multi-source person re-identification: a domain-specific adaptive framework. Pattern Recogn. 140, 109546 (2023)
S, S.R., Prasad, M.V.N.K., Balakrishnan, R.: Generative segment-pose representation based augmentation (GSRA) for unsupervised person re-identification. Image Vis. Comput. 131, 104632 (2023). https://doi.org/10.1016/j.imavis.2023.104632
Song, L., et al.: Unsupervised domain adaptive re-identification: theory and practice. Patt. Rec. 102, 107173 (2020). https://doi.org/10.1016/j.patcog.2019.107173
Tian, J., Teng, Z., Zhang, B., Wang, Y., Fan, J.: Imitating targets from all sides: an unsupervised transfer learning method for person re-identification. Int. J. Mach. Learn. Cybern. 12(8), 2281–2295 (2021). https://doi.org/10.1007/s13042-021-01308-6
Verma, A., Subramanyam, A.V., Wang, Z., Satoh, S., Shah, R.R.: Unsupervised domain adaptation for person re-identification via individual-preserving and environmental-switching cyclic generation. IEEE Trans. Multim. 25, 364–377 (2023). https://doi.org/10.1109/TMM.2021.3126404
Wang, M., Chen, J., Liu, H.: A novel multi-scale architecture driven by decoupled semantic attention transfer for person image generation. Comput. Graph. (2023)
Wang, W., Liao, S., Zhao, F., Kang, C., Shao, L.: Domainmix: learning generalizable person re-identification without human annotations. In: BMVC, p. 355 (2021)
Wang, Y., Liao, S., Shao, L.: Surpassing real-world source training data: random 3D characters for generalizable person re-identification. In: ACM MM, pp. 3422–3430 (2020). https://doi.org/10.1145/3394171.3413815
Wei, L., et al.: Person transfer GAN to bridge domain gap for person re-identification. In: CVPR, pp. 79–88 (2018). https://doi.org/10.1109/CVPR.2018.00016
Ye, M., et al.: Deep learning for person re-identification: a survey and outlook. IEEE Trans. PAMI 44(06), 2872–2893 (2022). https://doi.org/10.1109/TPAMI.2021.3054775
Yu, H., Wu, A., Zheng, W.: Cross-view asymmetric metric learning for unsupervised person re-identification. In: ICCV, pp. 994–1002 (2017). https://doi.org/10.1109/ICCV.2017.113
Zhang, C., et al.: Improving domain-adaptive person re-identification by dual-alignment learning with camera-aware image generation. IEEE T-CSVT 1, 4334–4346 (2020). https://doi.org/10.1109/TCSVT.2020.3047095
Zhang, L., Jiang, N., Diao, Q., Zhou, Z., Wu, W.: Person re-identification with pose variation aware data augmentation. Neural Comput. Appl. 34(14), 11817–11830 (2022). https://doi.org/10.1007/s00521-022-07071-1
Zhang, S., Hu, H.: Unsupervised person re-identification using unified domanial learning. Neural Process. Lett. 1–19 (2023)
Zheng, L., et al.: Scalable person re-identification: a benchmark. In: ICCV, pp. 1116–1124 (2015). https://doi.org/10.1109/ICCV.2015.133
Zhong, Z., et al.: CamStyle: a novel data augmentation method for person re-identification. IEEE T-IP 28(3), 1176–1190 (2019). https://doi.org/10.1109/TIP.2018.2874313
Zhou, K., et al.: Learning generalisable omni-scale representations for person re-identification. IEEE T-PAMI 1, 1–1 (2019). https://doi.org/10.1109/TPAMI.2021.3069237
Zhou, K., Liu, Z., Qiao, Y., Xiang, T., Loy, C.C.: Domain generalization: a survey. IEEE T-PAMI 45(4), 4396–4415 (2023). https://doi.org/10.1109/TPAMI.2022.3195549
Zhou, S., Wang, Y., Zhang, F., Wu, J.: Cross-view similarity exploration for unsupervised cross-domain person re-identification. Neural Comput. Appl. 33(9), 4001–4011 (2021). https://doi.org/10.1007/s00521-020-05566-3
Zou, Y., Yang, X., Yu, Z., Kumar, B.V.K.V., Kautz, J.: Joint disentangling and adaptation for cross-domain person re-identification. In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12347, pp. 87–104. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58536-5_6
Acknowledgements
Supported by the projects: “Law Enforcement agencies human factor methods and Toolkit for the Security and protection of CROWDs in mass gatherings” (LETSCROWD), EU Horizon 2020 programme, grant agreement No. 740466; “IMaging MAnagement Guidelines and Informatics Network for law enforcement Agencies” (IMMAGINA), European Space Agency, ARTES Integrated Applications Promotion Programme, contract No. 4000133110/20/NL/AF.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Putzu, L., Loddo, A., Delussu, R., Fumera, G. (2023). Specialise to Generalise: The Person Re-identification Case. In: Foresti, G.L., Fusiello, A., Hancock, E. (eds) Image Analysis and Processing – ICIAP 2023. ICIAP 2023. Lecture Notes in Computer Science, vol 14234. Springer, Cham. https://doi.org/10.1007/978-3-031-43153-1_32
Download citation
DOI: https://doi.org/10.1007/978-3-031-43153-1_32
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-43152-4
Online ISBN: 978-3-031-43153-1
eBook Packages: Computer ScienceComputer Science (R0)