Abstract
This paper provided a unified perspective to improve triplet loss function and proposed a substantial triplet loss named Penalty Metric Space Triplet Loss (Tri-PMS) based on this perspective for person re-identification. Specifically, we performed two kinds of reversed penalty on the positive and negative distance of triplet loss with hard mining (Tri-hard) and adaptively weakened the original margin based on penalty intensity. We have divided penalty metric space into two categories, including consistency and autonomy space. In consistency space, all batches shared the same penalty intensity without extra samples' information. Tri-PMS with consistency space is a tiny, friendly, and relatively efficient method for person re-identification tasks. In autonomy space, each batch holds their penalty based on triplet distribution. Tri-PMS with autonomy space grasps more information on each batch sample and learns a more perfect embedded space than consistency space. Our proposed method has obtained excellent performance on Market1501 with 94.8% mAP and 96.2% Rank-1, on DukeMTMC-reID with 89.7% mAP and 91.8% Rank-1. For CUHK03 and MSMT17, Tri-PMS also shows strong performance compared to other loss functions. More notably, Tri-PMS with autonomy space has stood the state-of-the-art performance on Occluded-DukeMTMC and Partial-iLIDS, two mainstream challenging datasets for person re-identification.
Similar content being viewed by others
References
Alemu L, Shah M, Pelillo M (2019) Deep Constrained Dominant Sets for Person Re-Identification, 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 2019, pp. 9854-9863, doi: https://doi.org/10.1109/ICCV.2019.00995
Chen W, Chen X (2017) Beyond triplet loss: a deep quadruplet network for person re-identification. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017:1320–1329. https://doi.org/10.1109/CVPR.2017.145
Chen J, Qin J, Yan Y, Huang L, Liu L, Zhu F, and Shao L (2020a) Deep Local Binary Coding for Person Re-Identification by Delving into the Details. In Proceedings of the 28th ACM International Conference on Multimedia (MM '20). Association for Computing Machinery, New York, NY, USA, 3034–3043. DOI:https://doi.org/10.1145/3394171.3413979
Chen Q, Zhang W, Fan J (2020b) Cluster-level Feature Alignment for Person Re-identification. arXiv:2008.06810
Dai Z, Chen M, Gu X, Zhu S, Tan P (2019) Batch dropblock network for person re-identification and beyond. 3690–3700. https://doi.org/10.1109/ICCV.2019.00379.
Deng J, Guo J (2019) ArcFace: additive angular margin loss for deep face recognition. IEEE/CVF Conf Comput vis Pattern Recogn (CVPR) 2019:4685–4694. https://doi.org/10.1109/CVPR.2019.00482
Fan X, Jiang W, Luo H, and Fei M (2018) SphereReID: deep hypersphere manifold embedding for person re-identification, https://doi.org/10.1016/j.jvcir.2019.01.010
Ge Y, Li Z, Zhao H, Yin G, Yi S, Wang X, Li H (2018) FD-GAN: pose-guided feature distilling GAN for robust person re-identification. In Proceedings of the 32nd International Conference on Neural Information Processing Systems. Curran Associates Inc., Red Hook, NY, USA, 1230–1241.
Guo J, Yuan Y, Huang L, Zhang C (2019) Beyond human parts: dual part-aligned representations for person re-identification. IEEE/CVF Int Conf Comput vis (ICCV) 2019:3641–3650. https://doi.org/10.1109/ICCV.2019.00374
Hadsell R, Chopra S, LeCun Y (2005) Learning a similarity metric discriminatively, with application to face verification. In 2013 IEEE Conference on Computer Vision and Pattern Recognition, San Diego, California, 2005 pp. 539–546.doi: https://doi.org/10.1109/CVPR.2005.202
Hao, L. (2019) Bag of tricks and a strong baseline for deep person re-identification. 1487–1495. https://doi.org/10.1109/CVPW.2019.00190.
He L, Liang J, Li H, et al (2018a) Deep spatial feature reconstruction for partial person re-identification: alignment-free approach.
He L, Sun Z, Zhu Y, et al (2018b) Recognizing partial biometric patterns.
He L, Wang Y, Liu W, et al (2019) Foreground-aware pyramid reconstruction for alignment-free occluded person re-identification.
Hermans A, Beyer L, Leibe B (2017) In defense of the triplet loss for person re-identifcation. https://arxiv.org/abs/1703.07737v2
Huang Z, Qin W, Luo F et al (2021) Combination of validity aggregation and multi-scale feature for person re-identification. J Ambient Intell Human Comput. https://doi.org/10.1007/s12652-021-03473-6
Huang H, Li D, Zhang Z, et al (2018) Adversarially occluded samples for person re-identification. 5098–5107. https://doi.org/10.1109/CVPR.2018.00535.
Jean P, Qin K, Liu G, Luo G, Agyemang B (2020) Enforcing affinity feature learning through self-attention for person re-identification. ACM Trans Multimed Comput Commun Appl 16:1–22. https://doi.org/10.1145/3377352
Jin X, Lan C, Zeng W (2020) Style Normalization and Restitution for Generalizable Person Re-Identification. IEEE/CVF Conf Comput vis Pattern Recogn (CVPR) 2020:3140–3149. https://doi.org/10.1109/CVPR42600.2020.00321
Liu H, Feng J, Qi M, Jiang J, Yan S (2017) End-to-end comparative attention networks for person re-identification. IEEE Trans Image Process 26(7):3492–3506. https://doi.org/10.1109/TIP.2017.2700762
Liu W & Wen Y& Yu Z, Yang M (2016) Large-margin softmax loss for convolutional neural networks. Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:507–516
Miao J, Wu Y, Liu P, et a (2019) Pose-guided feature alignment for occluded person re-identification. 542–551. https://doi.org/10.1109/ICCV.2019.00063.
Miao J, Wu Y, Yang Y (2021) identifying visible parts via pose estimation for occluded person re-identification
Quan R, Dong X, Wu Y, et al (2019) Auto-ReID: searching for a part-aware convnet for person re-identification, 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 2019, pp. 3749–3758, doi: https://doi.org/10.1109/ICCV.2019.00385.
Quispe R, Pedrini H (2020) Top-DB-net: top dropblock for activation enhancement in person re-identification
Ristani E, Solera F, Zou R, Cucchiara R, Tomasi C (2016) Performance measures and a data set for multi-target, multi-camera tracking
Schroff L, Kalenichenko D, Philbin J.(2015) Facenet: A unified embedding for face recognition and clustering. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston, MA: IEEE, 2015: 815–823
Selvaraju R, Cogswell M, Das A, et al (2017) Grad-cam: visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE International Conference on Computer Vision (CVPR). Venice, Italy: IEEE, 2017: 618–626
Sohn K (2016) Improved deep metric learning with multi-class N-pair loss objective. In Proceedings of the 30th International Conference on Neural Information Processing Systems. Curran Associates Inc., Red Hook, NY, USA, 1857–1865
Suh Y, Wang J, Tang S, Mei T, Lee KM (2018) Part-aligned bilinear representations for person re-identification. In: Ferrari V, Hebert M, Sminchisescu C, Weiss Y (eds) Computer vision – ECCV 2018. ECCV 2018. Lecture Notes in Computer Science, vol 11218. Springer, Cham. https://doi.org/10.1007/978-3-030-01264-9_25
Sun Y, Zheng L, Yang Y et al (2017) (2017) Beyond Part Models: Person Retrieval with Refined Part Pooling (and A Strong Convolutional Baseline). Springer, Cham
Sun Y, Cheng C, Zhang Y, Zhang C, Zheng L, Wang Z, Wei Y (2020a) Circle loss: a unified perspective of pair similarity optimization
Sun Y, Xu Q, Li Y, et al (2020b) Perceive where to focus: learning visibility-aware part-level features for partial person reidentification
Tay C P, Roy S, Yap K H (2020) AANet: Attribute Attention Network for Person Re-Identifications[C]// 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2020.
Van der Maaten L, Hinton G (2008) (2008) Visualizing data using t-SNE. J Mach Learn Res 9:11
Varior RR, Haloi M, Wang G (2016) Gated siamese convolutional neural network architecture for human re-identifcation. In: Leibe B, Matas J, Sebe N, Welling M (eds) Computer vision – ECCV 2016. ECCV 2016. Lecture Notes in computer science, vol 9912. Springer, Cham. https://doi.org/10.1007/978-3-319-46484-8_48
Wang F, Cheng J, Liu W, Liu H (2018) Additive margin softmax for face verification. IEEE Sign Process Lett 25(7):926–930. https://doi.org/10.1109/LSP.2018.2822810
Wang H et al (2018a) CosFace: large margin cosine loss for deep face recognition. IEEE/CVF Conf Comput vis Pattern Recogn 2018:5265–5274. https://doi.org/10.1109/CVPR.2018.00552
Wang G et al (2020) High-order information matters: learning relation and topology for occluded person re-identification. IEEE/CVF Conf Comput vis Pattern Recogn 2020:6448–6457. https://doi.org/10.1109/CVPR42600.2020.00648
Wang C, Pan ZG, Li XZ (2021) Multilevel metric rank match for person re-identification. Cogn Syst Res 65:98–106
Wang G, Yuan Y, Chen X, et al (2018b) Learning Discriminative Features with Multiple Granularities for Person Re-Identification. In Proceedings of the 26th ACM international conference on Multimedia. Association for Computing Machinery, New York, NY, USA, 274–282. DOI:https://doi.org/10.1145/3240508.3240552
Wei L, Zhang S, Wen G, et al (2017) Person transfer GAN to bridge domain gap for person re-identification
Xiao Q, Luo H, and Zhang C (2017) Margin sample mining loss: a deep learning based method for person re-identification, arXiv:1710.00478
Xuan Z, Hao L, Xing F, et al (2017) AlignedReID: surpassing human-level performance in person re-identification
Zheng W S, Gong S, Tao X (2011) Person re-identification by probabilistic relative distance comparison. In Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society, USA, 649–656. DOI:https://doi.org/10.1109/CVPR.2011.5995598
Zheng L, Shen L, Tian L, Wang S, Wang J, and Tian Q (2015) Scalable Person Re-identification: A Benchmark
Zheng Z, Yang X, Yu Z, Zheng L, Yang Y, and Kautz J (2020) Joint discriminative and generative learning for person re-identification
Zhong Z, Zheng L, Kang G, et al (2017a) Random erasing data augmentation
Zhong Z, Liang Z, D Cao, et al (2017b) Re-ranking person re-identification with k-reciprocal encoding
Zhong Z, Zheng L, Kang G, et al (2017c) Random Erasing Data Augmentation. Proceedings of the AAAI Conference on Artificial Intelligence, 2017c, 34(7)
Acknowledgements
This work is supported by National Key R&D Program of China (No. 2021YFC2009200).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Qin, Wc., Huang, Zy., Guan, Th. et al. Triplet penalty matters: penalty metric space triplet loss for person re-identification. J Ambient Intell Human Comput 14, 14029–14043 (2023). https://doi.org/10.1007/s12652-022-04109-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12652-022-04109-z