逐点特征匹配的跨域行人重识别方法

doi:10.16451/j.cnki.issn1003-6059.202206004

摘要
图/表
参考文献
相关文章 (9)

全文: PDF (1393 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对现有的直接跨数据集的行人重识别方法泛化性不足、跨域能力较差的问题,文中提出逐点特征匹配的跨域行人重识别方法,只需在源域上进行模型训练,在目标域上进行测试,就可达到较好效果.首先,为了解决网络对于跨域的行人图像风格、颜色等鲁棒性不高的问题,在ResNet50基础网络中引入实例归一化层,提取图像特征.然后,利用Transformer的多头自注意力模块与卷积结合,增强特征的表示能力.最后,通过在深层特征中建立一种逐点的特征映射关系,将图像匹配视为逐点寻找局部最优的过程,在未知场景中提升模型的抗视角变化能力,增强模型的泛化性.实验表明,文中方法在提高模型泛化能力上具有一定的优越性.

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	杨萍
	吴晓红
	何小海
	陈洪刚
	刘强
	李波

关键词 ：行人重识别, 跨域行人重识别, 实例归一化, 多头自注意力, 逐点匹配

Abstract：To improve the poor generalization and cross-domain capability of the existing direct cross-dataset person re-identification methods, a cross-domain person re-identification method based on point-by-point feature matching is proposed. By the proposed method, the model only needs to be trained on the source domain and tested on the target domain to achieve better results. Firstly, to improve the poor robustness of the network for style and color of cross-domain pedestrian images, instance normalization layer(IN) is introduced into the ResNet50 basic network to extract image features. Secondly, the multi-head self-attention module of Transformer is combined with convolution to enhance the representation ability of features. Finally, by establishing a point-by-point feature mapping relationship in the deep features, image matching is regarded as a point-by-point process of finding the local optimum to improve the ability of the proposed model to resist perspective changes in unknown scenes and enhance its generalization. The experimental results show that the advantages of the proposed method in improving the generalization ability.

Key words： Person Re-identification Cross-Domain Person Re-identification Instance Normalization Multi-head Self-Attention Point-by-Point Matching

收稿日期: 2022-02-25

ZTFLH:

TP 391

基金资助:国家自然科学基金项目(No.61871278)、四川省自然科学基金项目(No.2022NSFSC0922)、四川省科技计划项目(No.2021YFS0239)资助

通讯作者: 何小海,博士,教授,主要研究方向为图像处理、模式识别、计算机视觉、图像压缩.E-mail:hxh@scu.edu.cn.

作者简介: 杨萍,硕士研究生,主要研究方向为深度学习、行人重识别.E-mail:1047798178@qq.com.
吴晓红,博士,副教授,主要研究方向为图像处理、模式识别、计算机视觉.E-mail:wxh@scu.edu.cn.
陈洪刚,博士,副研究员,主要研究方向为图像/视频处理、计算机视觉、人工智能.E-mail:honggang_chen@scu.edu.cn.
刘强,博士研究生,主要研究方向为图像处理、行人重识别、计算机视觉.E-mail:liuliu408@163.com.
李波,硕士研究生,主要研究方向为计算机视觉、模式识别.E-mail:804463592@qq.com.

引用本文:

杨萍, 吴晓红, 何小海, 陈洪刚, 刘强, 李波. 逐点特征匹配的跨域行人重识别方法[J]. 模式识别与人工智能, 2022, 35(6): 516-525. YANG Ping, WU Xiaohong, HE Xiaohai, CHEN Honggang, LIU Qiang, LI Bo. Cross-Domain Person Re-identification Method Based on Point-by-Point Feature Matching. Pattern Recognition and Artificial Intelligence, 2022, 35(6): 516-525.

链接本文:

http://manu46.magtech.com.cn/Jweb_prai/CN/10.16451/j.cnki.issn1003-6059.202206004 或 http://manu46.magtech.com.cn/Jweb_prai/CN/Y2022/V35/I6/516

[1] GHEISSARI N, SEBASTIAN T B, HARTLEY R. Person Reidentification Using Spatiotemporal Appearance // Proc of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2006: 1528-1535.
[2] YU H X, WU A C, ZHENG W S. Cross-View Asymmetric Metric Learning for Unsupervised Person Re-identification // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2017: 994-1002.
[3] GE Y X, CHEN D P, LI H S.Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification[C/OL]. [2022-01-25].https://arxiv.org/pdf/2001.01526.pdf.
[4] ZHAI Y P, LU S J, YE Q X, et al. AD-Cluster: Augmented Discriminative Clustering for Domain Adaptive Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pa-ttern Recognition. Washington, USA: IEEE, 2020: 9018-9027.
[5] DENG W J, ZHENG L, YE Q X, et al. Image-Image Domain Ada-ptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 994-1003.
[6] ZHONG Z, ZHENG L, LI S Z, et al. Generalizing a Person Retrieval Model Hetero-and Homogeneously // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 176-192.
[7] CHOI Y, CHOI M, KIM M, et al. StarGAN: Unified Generative Adversarial Networks for Multi-domain Image-to-Image Translation // Proc of the IEEE/CVF Conference on Computer Vision and Pa-ttern Recognition. Washington, USA: IEEE, 2018: 8789-8797.
[8] LIU Q, HE X H, ZHANG M Z, et al. Feature Separation and Double Causal Comparison Loss for Visible and Infrared Person Re-identification. Knowledge-Based Systems, 2022, 239. DOI: 10.1016/j.knosys.2021.108042.
[9] MUANDET K, BALDUZZI D, SCHÖLKOPF B. Domain Generalization via Invariant Feature Representation // Proc of the 30th International Conference on Machine Learning. San Diego, USA: JMLR, 2013: 10-18.
[10] SHANKAR S, PIRATLA V, CHAKRABARTI S, et al. Generalizing Across Domains via Cross-Gradient Training[C/OL].[2022-01-25]. https://arxiv.org/pdf/1804.10745.pdf.
[11] JIN X, LAN C L, ZENG W J, et al. Style Normalization and Restitution for Generalizable Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2020: 3140-3149.
[12] SARFRAZ M S, SCHUMANN A, EBERLE A, et al. A Pose-Sensitive Embedding for Person Re-identification with Expanded Cross Neighborhood Re-ranking // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 420-429.
[13] ZHENG Z D, YANG X D, YU Z D, et al. Joint Discriminative and Generative Learning for Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2019: 2133-2142.
[14] XU J, ZHAO R, ZHU F, et al. Attention-Aware Compositional Network for Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 2119-2128.
[15] LI W, ZHU X T, GONG S G. Harmonious Attention Network for Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 2285-2294.
[16] CHEN G Y, LIN C Z, REN L L, et al. Self-Critical Attention Learning for Person Re-identification // Proc of the IEEE/CVF International Conference on Computer Vision. Washington, USA: IEEE, 2019: 9636-9645.
[17] ZHENG M, KARANAM S, WU Z Y, et al. Re-identification with Consistent Attentive Siamese Networks // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2019: 5728-5737.
[18] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An Image Is Worth 16×16 Words: Transformers for Image Recognition at Scale[C/OL].[2022-01-25]. https://arxiv.org/pdf/2010.11929.pdf.
[19] CARION N, MASSA F, SYNNAEVE G, et al. End-to-End Object Detection with Transformers // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2020: 213-229.
[20] HE S T, LUO H, WANG P C, et al. TransReID: Transformer-Based Object Re-identification // Proc of the IEEE/CVF International Conference on Computer Vision. Washington, USA: IEEE, 2021: 14993-15002.
[21] HUANG X, BELONGIE S. Arbitrary Style Transfer in Real-Time with Adaptive Instance Normalization // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2017: 1510-1519.
[22] PAN X G, LUO P, SHI J P, et al. Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 484-500.
[23] WANG H, WANG Y T, ZHOU Z, et al. CosFace: Large Margin Cosine Loss for Deep Face Recognition // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 5265-5274.
[24] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal Loss for Dense Object Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(2): 318-327.
[25] ZHONG Z, ZHENG L, CAO D L, et al. Re-ranking Person Re-identification with k-Reciprocal Encoding // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2017: 3652-3661.
[26] LIAO S C, SHAO L. Interpretable and Generalizable Person Re-identification with Query-Adaptive Convolution and Temporal Lifting // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2020: 456-474.
[27] WANG J Y, ZHU X T, GONG S G, et al. Transferable Joint Attri-bute-Identity Deep Learning for Unsupervised Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2018: 2275-2284.
[28] YANG Q Z, YU H X, WU A C, et al. Patch-Based Discriminative Feature Learning for Unsupervised Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Re-cognition. Washington, USA: IEEE, 2019: 3628-3637.
[29] ZHONG Z, ZHENG L, LUO Z M, et al. Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-identification // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2019: 598-607.
[30] MEKHAZNI D, BHUIYAN A, EKLADIOUS G, et al. Unsupervised Domain Adaptation in the Dissimilarity Space for Person Re-identification // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2020: 159-174.
[31] CHEN X D, LIU X C, LIU W, et al. Explainable Person Re-identification with Attribute-Guided Metric Distillation // Proc of the IEEE/CVF International Conference on Computer Vision. Wa-shington, USA: IEEE, 2021: 11793-11802.
[32] ZHOU K Y, YANG Y X, CAVALLARO A, et al. Learning Generalisable Omni-Scale Representations for Person Re-identification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021. DOI: 10.1109/TPAMI.2021.3069237.