计算机科学 ›› 2023, Vol. 50 ›› Issue (7): 160-166.doi: 10.11896/jsjkx.220600153
代雪松, 李小红, 张晶晶, 齐美彬, 刘一敏
DAI Xuesong, LI Xiaohong, ZHANG Jingjing, QI Meibin, LIU Yimin
摘要: 现有的无监督域自适应行人重新识别方法大多是将基于聚类的伪标签预测与特征微调相结合。由于域间存在差异,聚类过程中产生了不正确的伪标签,使伪标签存在一定的不可靠性,误导特征表示学习,从而影响域自适应模型的性能。基于此,首先设计一个新颖的基于反事实注意力学习的无监督域自适应网络,通过衡量注意力学习的质量对训练过程进行指导优化,促使模型关注更加精准的注意力特征,减少噪声伪标签的生成;其次提出了一种基于不确定性评估的噪声样本优化方法,通过测量基于平均教师方法的学生模型和教师模型输出特征之间的不一致性水平,将其作为目标域行人样本的不确定性分布,进而利用样本的不确定性对网络总体损失的各个部分进行合理加权,修正具有高不确定性的样本对模型总体损失的错误影响,进一步提升目标域的识别性能。实验数据表明,所提方法在源域DukeMTMC-reID/Market-1501和目标域Market-1501/DukeMTMC-reID上的实验结果都有显著提高,mAP和Rank-1分别达到了82.9%,93.6%和71.8%,84.4%。
中图分类号:
[1]FAN H,ZHENG L,YAN C,et al.Unsupervised person re-identification:Clustering and finetuning[J].ACM Transactions on Multimedia Computing,Communications,and Applications(TOMM),2018,14(4):1-18. [2]FU Y,WEI Y,WANG G,et al.Self-similarity grouping:A simple unsupervised cross domain adaptation approach for person re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:6112-6121. [3]ZHANG X,CAO J,SHEN C,et al.Self-Training with progressive augmentation for unsupervised cross-domain person re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:8222-8231. [4]JIN X,HE T,YIN Z,et al.Meta clustering learning for large-scale unsupervised person re-identification[J].arXiv:2111,10032,2021. [5]GE Y,CHEN D,LI H,et al.Mutual mean-teaching:Pseudo label refinery for unsupervised domain adaptation on person re-identification[J].arXiv:2001.01526,2020. [6]ZHAO F,LIAO S,XIE G,et al.Unsupervised domain adaptation with noise resistible mutual-training for person re-identification[C]//European Conference on Computer Vision.Springer,2020:526-544. [7]HE K,FAN H,WU Y,et al.Momentum contrast for unsuper-vised visual representation learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:9729-9738. [8]ZHONG Z,ZHENG L,LUO Z,et al.Invariance matters:Exemplar memory for domain adaptive person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:598-607. [9]LIU J,ZHA Z J,CHEN D,et al.Adaptive transfer network for cross-domain person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:7202-7211. [10]SONG L,WANG C,ZHANG L,et al.Unsupervised domainadaptive re-identification:Theory and practice[J].Pattern Recognition,2020,102:107173. [11]ZHAI Y,LU S,YE Q,et al.Ad-cluster:Augmented discriminative clustering for domain adaptive person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:9021-9030. [12]ZHENG K,LAN C,ZENG W,et al.Exploiting sample uncertainty for domain adaptive person re-identification[J].arXiv:2012.08733,2020. [13]ZHENG K,LIU W,HE L,et al.Group-aware label transfer for domain adaptive person re-identification[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:5310-5319. [14]PEARL J.Direct and indirect effects[M].Probabilistic andCausal Inference:The Works of Judea Pearl.2022:373-392. [15]VANDERWEELE T.Explanation in causal inference:Developments in mediation and interaction[J].International Journal of Epidemiology,2016,45(6):1904-1908. [16]KENDALL A,GAL Y.What uncertainties do we need in baye-sian deep learning for computer vision?[J].arXiv:1703.04977,2017. [17]RAO Y,CHEN G,LU J,et al.Counterfactual attention learning for fine-grained visual categorization and re-identification[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2021:1025-1034. [18]VANDERWEELE T.Explanation in causal inference:methodsfor mediation and interaction[M].NewYork:Oxford University Press,2015. [19]ZHENG L,SHEN L,TIAN L,et al.Scalable person re-identification:A benchmark[C]//Proceedings of the IEEE Interna-tional Conference on Computer Vision.2015:1116-1124. [20]RISTANI E,SOLERA F,ZOU R,et al.Performance measures and a data set for multi-target,multi-camera tracking[C]//European Conference on Computer Vision.Springer,2016:17-35. [21]LIN Y,DONG X,ZHENG L,et al.A bottom-up clustering approach to unsupervised person re-identification[C]//Procee-dings of the AAAI Conference on Artificial Intelligence.2019:8738-8745. [22]YU H,ZHENG W,WU A,et al.Unsupervised person re-identification by soft multilabel learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:2148-2157. [23]LI Y,LIN C,LIN Y,et al.Cross-dataset person re-identification via unsupervised pose disentanglement and adaptation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2019:7919-7929. [24]WANG D,ZHANG S.Unsupervised person re-identification via multi-label classification[C] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:10981-10990. [25]HUANG Y,PENG P,JIN Y,et al.Domain adaptiveattention model for unsupervised cross-domain person re-identification[J].arXiv:1905.10529,2019. |
|