时间特征互补的无监督视频行人重识别

doi:10.19678/j.issn.1000-3428.0062643

摘要/Abstract

摘要： 目前的视频行人重识别方法不能有效提取视频帧之间的时空信息，且需要解决人工标签的问题，提出一种时间特征互补的无监督视频行人重识别方法。利用时间特征擦除网络模块对视频帧与帧之间的时间信息特征及空间信息特征进行擦除提取，挖掘行人不同的特征以减少每帧特征的冗余，进而得到目标行人不同视觉的完整特征。通过约束性无监督层次聚类模块计算每个样本之间的距离得到高质量的不同身份集群，根据集群之间距离进行聚类生成高质量的伪标签，提高不同身份极度相似的样本识别性，并根据PK抽样困难样本三元组损失模块从已经聚类好的结果中抽取样本生成一个新的数据集，以便在每次聚类迭代后进行训练，减少困难样例对模型的影响。在MARS数据集和DukeMTMC-VideoReID数据集上的实验结果表明，该方法的平均精度均值分别达到了46.4%和72.5%，Rank-1分别达到了69.3%和80.5%，性能指标优于传统的RACM和DAL等方法。

关键词: 无监督视频行人重识别, 离散度, 聚类, 多样性约束, 时间特征

Abstract: Current video person re-identification methods cannot effectively extract spatiotemporal information between video frames, and the manual labeling problem must be solved.Therefore, this article proposes an unsupervised video person re-identification method with complementary temporal features.The Temporal Feature Erasure(TFE) network module extracts temporal and spatial information features between video frames while mining different features of pedestrians and reducing the redundancy of features in each frame, thus obtaining the complete features of different visuals of the target pedestrian.The constrained unsupervised hierarchical clustering module obtains high-quality distinct identity clusters by calculating the distance between samples.Clustering based on the distance between clusters is used to generate high-quality pseudo-labels for improving the recognition of extremely similar distinct identity samples.PK sampling of difficult-sample triad loss generates a new dataset by extracting samples from the already clustered results to perform training after each clustering iteration.This reduces the impact of difficult samples on the model.The method is evaluated experimentally on the Motion Analysis and Re-identification Set(MARS) and Duke Multi-Tracking Multi-Camera Video-based Re-IDentification (DukeMTMC-VideoReID) datasets, with a mean Average Precision (mAP) reaching 46.4% and 72.5%, respectively, and Rank-1 reaching 69.3% and 80.5%, respectively.The performance index is superior to those of the traditional RACM and DAL methods.

Key words: unsupervised video person re-identification, dispersion, clustering, diversity constraint, temporal feature

中图分类号:

TP391

王福银, 韩华, 黄丽, 陈益平. 时间特征互补的无监督视频行人重识别[J]. 计算机工程, 2022, 48(10): 313-320.

WANG Fuyin, HAN Hua, HUANG Li, CHEN Yiping. Unsupervised Video Person Re-identification with Complementary Temporal Features[J]. Computer Engineering, 2022, 48(10): 313-320.

https://www.ecice06.com/CN/Y2022/V48/I10/313

图/表 11

参考文献

[1] 董亚超, 刘宏哲, 徐成.基于显著性多尺度特征协作融合的行人重识别方法[J].计算机工程, 2021, 47(6):234-244, 252. DONG Y C, LIU H Z, XU C.Person re-identification method based on joint fusion of saliency multi-scale features[J].Computer Engineering, 2021, 47(6):234-244, 252.(in Chinese)
[2] 姚蓝, 兰巨龙, 胡涛.基于聚类优化的SDN多域自适应管理方法[J].计算机工程, 2019, 45(6):119-126. YAO L, LAN J L, HU T.Adaptive management method for SDN multi-domain based on clustering optimization[J].Computer Engineering, 2019, 45(6):119-126.(in Chinese)
[3] HAN H, ZHOU M C, SHANG X W, et al.KISS for rapid and accurate pedestrian re-identification[J].IEEE Transactions on Intelligent Transportation Systems, 2021, 22(1):394-403.
[4] 田智慧, 郑付科, 高需.内容一致性行人重识别算法[J].计算机工程, 2021, 47(3):237-242. TIAN Z H, ZHENG F K, GAO X.Content-consistent pedestrian re-identification algorithm[J].Computer Engineering, 2021, 47(3):237-242.(in Chinese)
[5] LIU H, FENG J S, QI M B, et al.End-to-end comparative attention networks for person re-identification[J].IEEE Transactions on Image Processing, 2017, 26(7):3492-3506.
[6] CHEN G Y, LU J W, YANG M, et al.Spatial-temporal attention-aware learning for video-based person re-identification[J].IEEE Transactions on Image Processing, 2019, 28(9):4192-4205.
[7] HAN H, ZHOU M C, ZHANG Y J.Can virtual samples solve small sample size problem of KISSME in pedestrian re-identification of smart transportation?[J].IEEE Transactions on Intelligent Transportation Systems, 2020, 21(9):3766-3776.
[8] DING G D, KHAN S, TANG Z M, et al.Towards better validity:dispersion based clustering for unsupervised person re-identification[EB/OL].[2021-08-01].https://arxiv.org/abs/1906.01308.
[9] LIN Y T, DONG X Y, LIANG Y Z, et al.A bottom-up clustering approach to unsupervised person re-identification[C]//Proceedings of AAAI Conference on Artificial Intelligence.[S.1.]:AAAI Press, 2019:549-562.
[10] GHIASI G, LIN T Y, LE Q V.DropBlock:a regularization method for convolutional networks[EB/OL].[2021-08-01].https://arxiv.org/abs/1810.12890.
[11] ZHENG L, BIE Z, SUN Y F, et al.Mars:video benchmark for large-scale person re-identification[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2016:868-884.
[12] ZHONG Z, ZHENG L, LUO Z M, et al.Invariance matters:exemplar memory for domain adaptive person re-identification[C]//Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2015:598-607.
[13] CHENG K Y, TAO F, ZHAN Y Z, et al.Hierarchical attributes learning for pedestrian re-identification via parallel stochastic gradient descent combined with momentum correction and adaptive learning rate[J].Neural Computing and Applications, 2020, 32(10):5695-5712.
[14] LIAO X Y, HE L X, YANG Z W, et al.Video-based person re-identification via 3D convolutional networks and non-local attention[C]//Proceedings of ACCVʼ18.Berlin, Germany:Springer, 2018:620-634.
[15] KIRAN M, BHUIYAN A, BLAIS-MORIN L A, et al.A flow-guided mutual attention network for video-based person re-identification[EB/OL].[2021-08-01].https://arxiv.org/abs/2008.03788.
[16] ZHENG L, BIE Z, SUN Y F, et al.MARS:a video benchmark for large-scale person re-identification[C]//Proceedings of ACCVʼ16.Berlin, Germany:Springer, 2016:868-884.
[17] ZHANG Z Z, LAN C L, ZENG W J, et al.Multi-granularity reference-aided attentive feature aggregation for video-based person re-identification[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2020:10404-10413.
[18] XIAO T, LI S, WANG B C, et al.Joint detection and identification feature learning for person search[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2017:3415-3424.
[19] YE M, LI J W, MA A J, et al.Dynamic graph co-matching for unsupervised video-based person re-identification[J].IEEE Transactions on Image Processing, 2019, 28(6):2976-2990.
[20] LIU Z M, WANG D, LU H C.Stepwise metric promotion for unsupervised video person re-identification[C]//Proceedings of 2017 IEEE International Conference on Computer Vision.Washington D.C., USA:IEEE Press, 2017:2448-2457.
[21] YE M, LAN X Y.Robust anchor embedding for unsupervised video person re-identification in the wild[C]//Proceedings of European Conference on Computer Vision.Berlin, Germany:Springer, 2018:170-186.
[22] CHEN Y B, ZHU X T, GONG S.Deep association learning for unsupervised video person re-identification[EB/OL].[2021-08-01].https://arxiv.org/pdf/1808.07301.pdf.
[23] WU Y, LIN Y T, DONG X Y, et al.Exploit the unknown gradually:one-shot video-based person re-identification by stepwise learning[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2018:5177-5186.
[24] LIN Y T, DONG X Y, ZHENG L, et al.A bottom-up clustering approach to unsupervised person re-identification[C]//Proceedings of AAAI Conference on Artificial Intelligence.Menlo Park, USA:AAAI Press, 2019:8738-8745.

选择文件类型/文献管理软件名称

选择包含的内容