ORITRIP: ENHANCING PERSON RE-IDENTIFICATION WITH ORIENTATION-AWARE TRIPLET LOSS
DOI:
https://doi.org/10.62985/j.huit_ojs.vol26.no2E.404Keywords:
Human body orientation, metric learning, person re-identification, triplet loss, viewpoint variationAbstract
Person re-identification faces real challenges when the same person appears in different poses and camera angles across multiple views. While standard triplet loss functions are effective for learning features that distinguish between individuals, they struggle with capturing spatial orientation details. This creates problems when the system needs to match the same person who appears at different angles or in different poses. This study presents an orientation-aware triplet loss function that incorporates human body orientation information during training. The approach builds on conventional triplet loss by adding orientation constraints that capture the geometric relationships between different poses and viewpoints of the same person. This orientation-aware mechanism can maintain consistent feature representations across varying body orientations while keeping the ability to distinguish between different individuals. Thorough evaluations on three popular person re-identification datasets showed performance gains when integrating the orientation-aware triplet loss with state-of-the-art algorithms. The results demonstrated better matching accuracy compared to methods using traditional triplet loss. In addition, the orientation-aware approach tackles key limitations in existing person re-identification systems by modeling the geometric properties of human subjects during feature learning. This research contributes to advancing person re-identification technology by providing a more effective loss function that better captures the inherent spatial characteristics of human appearance across different orientations.
References
[1] L. Zheng, Y. Yang, and A. G. Hauptmann, "Person Re-identification: Past, Present and Future," arXiv:1610.02984, 2016.
[2] M. Ye, J. Shen, G. Lin, T. Xiang, L. Shao and S. C. H. Hoi, "Deep Learning for Person Re-Identification: A Survey and Outlook," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 6, pp. 2872-2893, 1 June 2022, doi: https://doi.org/10.1109/TPAMI.2021.3054775.
[3] J. Liu, Z.-J. Zha, H. Xie, Z. Xiong, and Y. Zhang, "CA3Net: Contextual-Attentional Attribute-Appearance Network for Person Re-Identification," in Proc. 26th ACM Int. Conf. Multimedia, Seoul, Republic of Korea, 2018, pp. 737-745, doi: https://doi.org/10.1145/3240508.3240585.
[4] Y. Yuan, W. Chen, Y. Yang and Z. Wang, "In Defense of the Triplet Loss Again: Learning Robust Person Re-Identification with Fast Approximated Triplet Loss and Label Distillation," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. Workshops (CVPRW), Seattle, WA, USA, 2020, pp. 1454-1463, doi: https://doi.org/10.1109/CVPRW50498.2020.00185.
[5] J. Miao, Y. Wu and Y. Yang, "Identifying Visible Parts via Pose Estimation for Occluded Person Re-Identification," IEEE Trans. Neural Netw. Learn. Syst., vol. 33, no. 9, pp. 4624-4634, Sept. 2022, doi: https://doi.org/10.1109/TNNLS.2021.3059515.
[6] Y. Zhai, X. Han, W. Ma, X. Gou and G. Xiao, "PGMANet:Pose-Guided Mixed Attention Network for Occluded Person Re-Identification," in Proc. Int. Joint Conf. Neural Netw. (IJCNN), Shenzhen, China, 2021, pp. 1-8, doi: https://doi.org/10.1109/IJCNN52387.2021.9534442.
[7] D. Chen, Zejian Yuan, G. Hua, N. Zheng and J. Wang, "Similarity learning on an explicit polynomial kernel feature map for person re-identification," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Boston, MA, USA, 2015, pp. 1565-1573, doi: https://doi.org/10.1109/CVPR.2015.7298764.
[8] Y. Zhang, B. Li, H. Lu, A. Irie and X. Ruan, "Sample-specific SVM learning for person re-identification," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Las Vegas, NV, USA, 2016, pp. 1278-1287, doi: https://doi.org/10.1109/CVPR.2016.143.
[9] T. Q. Nguyen, O. D. A. Prima, and K. Hotta, "CORE-ReID: Comprehensive optimization and refinement through ensemble fusion in domain adaptation for person re-identification," Software, vol. 3, no. 2, pp. 227–249, 2024, doi: https://doi.org/10.3390/software3020012.
[10] T. Q. Nguyen, O. D. A. Prima, S. A. Irfan, H. D. Purnomo, and R. Tanone, "CORE-ReID V2: Advancing the domain adaptation for object re-identification with optimized training and ensemble fusion," AI Sens, vol. 1, no. 4, pp. 432–455, 2025, doi: https://doi.org/10.3390/aisens1010004.
[11] V. Somers, C. D. Vleeschouwer and A. Alahi, "Body Part-Based Representation Learning for Occluded Person Re-Identification," in Proc. IEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), Waikoloa, HI, USA, 2023, pp. 1613-1623, doi: https://doi.org/10.1109/WACV56688.2023.00166.
[12] A. Ghosh, K. Shanmugalingam and W. -Y. Lin, "Relation Preserving Triplet Mining for Stabilising the Triplet Loss in Re-identification Systems," in Proc. IEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), Waikoloa, HI, USA, 2023, pp. 4829-4838, doi: https://doi.org/10.1109/WACV56688.2023.00482.
[13] M. Kim, M. Cho, H. Lee, S. Cho and S. Lee, "Occluded Person Re-Identification Via Relational Adaptive Feature Correction Learning," in Proc. IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP), Singapore, Singapore, 2022, pp. 2719-2723, doi: https://doi.org/10.1109/ICASSP43922.2022.9746734.
[14] S. He, H. Luo, P. Wang, F. Wang, H. Li and W. Jiang, "TransReID: Transformer-based object re-identification," in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), Montreal, QC, Canada, 2021, pp. 14993-15002, doi: https://doi.org/10.1109/ICCV48922.2021.01474.
[15] M. Sun, J. Ren and Y. Zhang, "EdgeVPR: Transformer-Based Real-Time Video Person Re-Identification at the Edge," in Proc. IEEE 44th Int. Conf. Distrib. Comput. Syst. (ICDCS), Jersey City, NJ, USA, 2024, pp. 13-24, doi: https://doi.org/10.1109/ICDCS60910.2024.00011.
[16] D. Yi, Z. Lei, S. Liao and S. Z. Li, "Deep Metric Learning for Person Re-identification," in Proc. 22nd Int. Conf. Pattern Recognit. (ICPR), Stockholm, Sweden, 2014, pp. 34-39, doi: https://doi.org/10.1109/ICPR.2014.16.
[17] G. Zhang, J. Li, and Z. Ye, "Unsupervised Joint Contrastive Learning for Aerial Person Re-Identification and Remote Sensing Image Classification," Remote Sens, vol. 16, no. 2, 2024, doi: https://doi.org/10.3390/rs16020422.
[18] W. Chen, X. Chen, J. Zhang and K. Huang, "Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-identification," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Honolulu, HI, USA, 2017, pp. 1320-1329, doi: https://doi.org/10.1109/CVPR.2017.145.
[19] Y. Sun, C. Cheng, Y. Zhang, C. Zhang, L. Zheng, and Z. Wang, "Circle Loss: A Unified Perspective of Pair Similarity Optimization," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Seattle, WA, USA, 2020, pp. 6397-6406, doi: https://doi.org/10.1109/CVPR42600.2020.00643.
[20] Z. Cao, G. Hidalgo, T. Simon, S. -E. Wei and Y. Sheikh, "OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields," IEEE Trans. Pattern Anal. Mach. Intell., vol. 43, no. 1, pp. 172-186, 1 Jan. 2021, doi: https://doi.org/10.1109/TPAMI.2019.2929257.
[21] K. Sun, B. Xiao, D. Liu and J. Wang, "Deep High-Resolution Representation Learning for Human Pose Estimation," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Long Beach, CA, USA, 2019, pp. 5686-5696, doi: https://doi.org/10.1109/CVPR.2019.00584.
[22] L. Ma, H. Liu, L. Hu, C. Wang, and Q. Sun, "Orientation Driven Bag of Appearances for Person Re-identification," ArXiv:1605.02464, 2016.
[23] S. Andavar and P. Elango, "Person re-identification based on pose angle estimation and multi-feature extraction," IET Biometrics, vol. 7, no. 4, pp. 365-370, 2018, doi: https://doi.org/10.1049/iet-bmt.2016.0198.
[24] X. Sun and L. Zheng, "Dissecting person re-identification from the viewpoint of viewpoint," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Long Beach, CA, USA, 2019, pp. 608-617, doi: https://doi.org/10.1109/CVPR.2019.00070.
[25] W. Chen, X. Xu, J. Jia, H. Luo, Y. Wang, and F. Wang, "Beyond appearance: A semantic controllable self-supervised learning framework for human-centric visual tasks," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Vancouver, BC, Canada, 2023, pp. 15050-15061, doi: https://doi.org/10.1109/CVPR52729.2023.01445.
[26] C. Wu, Y. Chen, J. Luo, C.-C. Su, A. Dawane, and B. Hanzra, "MEBOW: Monocular Estimation of Body Orientation in the Wild," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Seattle, WA, USA, 2020, pp. 3448-3458, doi: https://doi.org/10.1109/CVPR42600.2020.00351.
[27] S. Li, L. Sun, and Q. Li, "CLIP-ReID: Exploiting Vision-Language Model for Image Re-identification without Concrete Text Labels," in Proc. AAAI Conf. Artif. Intell., 2023, pp. 1405-1413, doi: https://doi.org/10.1609/aaai.v37i1.25225.
[28] L. He, X. Liao, W. Liu, X. Liu, P. Cheng, and T. Mei, "FastReID: A Pytorch Toolbox for General Instance Re-identification," in Proc. 31st ACM Int. Conf. Multimedia, New York, NY, USA, 2023, pp. 9664–9667, doi: https://doi.org/10.1145/3581783.3613460.
[29] L. Zheng, L. Shen, L. Tian, S. Wang, J. Wang and Q. Tian, "Scalable Person Re-identification: A Benchmark," in Proc. IEEE Int. Conf. Comput. Vis. (ICCV), Santiago, Chile, 2015, pp. 1116-1124, doi: https://doi.org/10.1109/ICCV.2015.133.
[30] W. Li, R. Zhao, T. Xiao and X. Wang, "DeepReID: Deep filter pairing neural network for person re-identification," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Columbus, OH, USA, 2014, pp. 152-159, doi: https://doi.org/10.1109/CVPR.2014.27.
[31] L. Wei, S. Zhang, W. Gao and Q. Tian, "Person transfer GAN to bridge domain gap for person re-identification," in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Salt Lake City, UT, USA, 2018, pp. 79-88, doi: https://doi.org/10.1109/CVPR.2018.00016.
