Multimedia ResearchISSN:2582-547X

Fusion of Attention and Graph-Based Deep Learning for Person Re-identification: An Advancing Multi-Camera Surveillance

Abstract

Person re-identification in multi-camera surveillance is an emerging research area with great potential. Existing methods face challenges in accurately identifying people due to their limitations in capturing complex features and modeling inter-camera relationships. To address these problems, we propose a novel approach that combines attention mechanisms and graph-based deep learning. Current methods struggle to capture the complex relationships between many individual functions and cameras. Our method fuses an end-to-end attention mechanism for feature extraction with an undirected graph model to capture inter-camera relationships. This fusion is supported by tailored loss functions and training processes. Tailored loss functions and a robust training process support this fusion. Experimentally evaluated on reference datasets such as Market-1501 and DukeMTMC-reID, our framework consistently outperforms existing methods, demonstrating a significant increase in precision. The proposed fusion approach represents a promising development in the field of multi- camera surveillance systems, as it effectively solves the problem and facilitates person re-identification.

References

  • W. Li, X. Zhu, and S. Gong, “Harmonious attention network for person reidentification,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., Jun. 2018, pp. 2285–2294.
  • Z. Zhang, C. Lan, W. Zeng, X. Jin, and Z. Chen, “Relation-aware global attention for Re-ID,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2020, pp. 3186–3195.
  • H. Liu, J. Feng, M. Qi, J. Jiang, and S. Yan, “End-to-end comparative attention networks for Re-ID,” IEEE TIP, vol. 27, no. 7, pp. 3492–3506, Jul. 2017.
  • B. Chen, W. Deng, and J. Hu, “Mixed high-order attention network for Re-ID,” in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), Oct. 2019, pp. 371–381.
  • J. Xu, R. Zhao, F. Zhu, H. Wang, and W. Ouyang, “Attention-aware compositional network for Re-ID,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., Jun. 2018.
  • Fu, J., Zheng, H., Mei, T., & Luo, J. (2019). Dual Attention Network for Scene Segmentation. CVPR, 3141–3150.
  • Liu, W., Wen, Y., Yu, Z., Li, M., Rajan, D., & Song, L. (2017). SphereFace: Deep Hypersphere Embedding for Face Recognition. CVPR, 212–220.
  • Ejaz Ahmed, Michael Jones, and Tim K Marks, “An improved deep learning architecture for Re-ID,” Differences, 5, 25, (2015).
  • Yimin Wang, Ruimin Hu, Chao Liang, Chunjie Zhang, and Qingming Leng, “Camera compensation using a feature projection matrix for person reidentification,” IEEE Trans. Circuits Syst. Video Technol., 24(8), 1350–1361, (2014).
  • Rui Zhao, Wanli Ouyang, and Xiaogang Wang, “Unsupervised salience learning for Re-ID,” in Computer Vision and Pattern Recognition (CVPR), 2013 IEEE Conf., pp. 3586–3593. IEEE, (2013).
  • Q. Zhang, S. Zhao, Y. Yang, F. Ji, and Y. Wang, "AGW: Attention Guided GCNs for Re-ID," IEEE Trans. Multimedia, vol. 22, no. 12, pp. 3169–3182, Dec. 2020.
  • Wei-Shi Zheng, Shaogang Gong, and Tao Xiang, “Person reidentification by probabilistic relative distance comparison,” in CVPR, 2011 IEEE Conf., pp. 649–656. IEEE.
  • Taiqing Wang, Shaogang Gong, Xiatian Zhu, and Shengjin Wang, “Person re-identification by discriminative selection in video ranking,” (2016).
  • Z. Zhang, J. Chen, and L. Jia, "Relation-Aware Global Attention for Person Re-Identification," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2020.
  • K. Zheng, Y. Wang, H. Liu, and H. Ye, "Graph-Based Re-ID with Attribute-Driven Attention," in Proc. Eur. Conf. Comput. Vis. (ECCV), 2020.
  • X. Wu, H. Liu, and S. Zhang, "GCNs for Text-based Re-ID," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2019.
  • H. Yan, H. Wu, X. Zhu, and S. Gong, "MuDeep: A Multi-Scale Graph Neural Network for Human-Object Interaction Detection in Multi-Camera Videos," in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), 2019.
  • X. Wu, Y. Yang, Q. Zhang, et al., "Attention in Convolutional LSTM for Gesture Recognition," in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), 2019.
  • A. Hermans, L. Beyer, and B. Leibe, “In defense of the triplet loss for Re-ID,” 2017, arXiv:1703.07737.
  • Y. Fu, C. Wei, Y. Wang, and Z. Xiong, "Dual Attention Matching Network for Context-Aware Feature Sequence Based Re-ID," in Proc. CVPR, pp. 7114–7123, 2019.
  • M. Jaderberg, K. Simonyan, A. Zisserman, et al., "Spatial Transformer Networks," in Adv. Neural Inf. Process. Syst. (NeurIPS), pp. 2017–2025, 2015.
  • X. Wang, Z. Li, Y. Gong, and J. Huang, "Unlabeled Samples Generated by GAN Improve the Re-ID Baseline in Vitro," in Proc. CVPR, pp. 3757–3765, 2018.
  • L. Chen, L. Fan, W. Xu, W. Wang, and S. Gong, "Fusion of Discriminative and Coherent Attention Maps for Re-ID," IEEE Access, vol. 8, pp. 48839–48848, 2020.
  • X. Qian, Y. Fu, T. Xiang, X. Wang, and W. Qiu, "Learning Spatial-Temporal Regularized Correlation Filters for Visual Tracking," CVPR, pp. 4905–4914, 2019.
  • S. Wang, S. Gong, X. Zhu, and W. Wang, "Re-ID by Video Ranking," CVPR, pp. 688–704, 2018.
  • Z. Zheng, L. Zheng, Y. Yang, and P. Yu, "Joint Discriminative and Generative Learning for Re-ID," CVPR, pp. 2138–2147, 2019.
  • Y. Yan, J. Ni, X. Liang, X. Zhang, and J. Tang, "GCNs for Temporal Action Localization," CVPR, pp. 4931–4940, 2020.
  • X. Liu, Z. Song, Y. Wu, et al., "A Simple and Strong Baseline for Universal Targeted Attacks on Re-ID," ICCV, pp. 2752–2761, 2019.
  • Y. Zhang, X. Zhu, Y. Li, and X. Huang, "Attention-Aware Composition Learning for Re-ID," CVPR, pp. 9681–9690, 2019.
  • Y. Yan, W. Wu, Y. Wang, and C. Qian, "Spatial-Temporal GCN for Video-based Re-ID," CVPR, pp. 12908–12917, 2020.
  • L. Wu, Y. Wang, W. Lin, and Z. Gao, "Unsupervised Cross-modality Re-identification via Graph-structured Semantic Transfer," CVPR, pp. 2171–2180, 2020.
  • X. Xie, Z. Song, X. Wu, et al., "Comprehensive Survey of Person Re-identification with Deep Learning," arXiv preprint arXiv:2001.04193, 2020.
  • Sharif, S., Lodhi, R. N., Iqbal, K., & Saddique, F. (2022). Gender disparity in leadership boosts affective commitment and tacit knowledge sharing about libraries. International Journal of Organizational Analysis, 30(5), 1212–1234.
  • Saddique, F., Patel, K. R., Niaz, M., Chukwu, M. U., & Nwagwu, U. (2023). Impact of Supply Chain Transformation on Supply Chain Performance: The Empirical Study that Bases on Mediating Role of Supply Chain Resilience on Construction Organization in Pakistan. Asian Journal of Engineering, Social and Health, 2(9), 1072–1086.
  • Saddique, F., Fatima, H., Muzammil, M. H., & Mushtaq, N. (2023). Impact of ethical leadership on creative work performance: the empirical mediating roles of work engagement and attitude toward performing wells in construction organizations in Pakistan. Journal of Social Sciences Review, 3(1), 607–622.
  • Mushtaq, N., Saddique, F., Abbas, Z., & Nawaz, M. (2023). How important is Efficiency in any Organization? “Estimating the Efficiency Reform of Power Distribution Companies in Punjab Province, Pakistan” (LESCO, FESCO & IESCO). Traditional Journal of Law and Social Sciences, 2(01), 44–59.
  • Saddique, F., Mushtaq, N., Abbas, Z., & Nawaz, M. (2023). Creative Performance of Healthcare Professionals (Nurses) in Hospitals Lahore, Pakistan. Traditional Journal of Law and Social Sciences, 2(01), 60–72.
  • Saddique, F., Nwagwu, U., Mushtaq, N., Lamiaa, B., & Ali, A. (2023). Implementation of digitalization supply chain helps in gaining of competitive advantages as mediating role in the supply chain performance in construction organization in Pakistan. Traditional Journal of Humanities, Management, and Linguistics, 2(01), 14–27.
  • Saddique, F., Usman, M., Nawaz, M., & Mushtaq, N. (2020). Entrepreneurial orientation and human resource management: The mediating role of Artificial Intelligence. Ilkogretim Online, 19(4), 4969–4978.
  • Saddique, F., Mushtaq, N., Mehmood, R. S., Durrani, M. Z., & Nawaz, M. (2021). The effects of information asymmetry, accounting information and personal values on investment satisfaction mediating by direct investment decision makers. Psychology and Education, 58(5), 4225–4252.
  • Saddique, F., Khan, K. N., Joyia, S. Z., & Jabeen, A. (2024). Impact of Internet of Things on Sustainable Supply Chain Performance: The Empirical Parallel Mediation Models. Pakistan Islamicus, 4(02), 76–89.
  • Saddique, F., Mushtaq, N., Hasan, R., Rehman, Z., & Ali, M. (2022). Access and Secure Storage Based Block Chain Scheme with IPFS Implemented in Electronic Medical Record in Lahore, Pakistan. Netw. Commun. Syst. JNACS, 5(4).
  • Nawaz, M., Khan, S. A., Rana, S. L., & Saddique, F. (2020). Role of Human Resource Best Practices to Improve Quality Education of Pakistani Schools. International Journal of Management (IJM), 11(11).
  • Saddique, F., Mushtaq, N., Imran, Z., Nawaz, M., & Usman, M. (2021). Impact of Green Supply Chain Management Practices on Environment Performance and Competitive Advantage.
    •

Access options

  • Price $
  • Offer $
Buy article PDF

DOI : https://doi.org/10.46253/j.mr.v8i2.a3

Author information

Affiliations

Wuhan Textile University shawnajhrk@gmail.com

Wuhan Textile University saharali99981@gmail.com

Wuhan Textile University
hkralfahd@gmail.com

Publisher Information

  • Received

    20 March,

  • Revised

    26 April

  • Accepted

    5 May

Keywords
Person Re-Identification, Multi-Camera Surveillance, Attention Mechanisms, Graph-Based Models, Feature Extraction, Loss Function, Accuracy Improvement.

Publisher : Resbee Info Technologies Pvt Ltd