典型文献
Domain Adaptive Learning with Multi-Granularity Features for Unsupervised Person Re-identification
文献摘要：
Unsupervised person re-identification(Re-ID)aims to improve the model's scalability and ob-tain better Re-ID results in the unlabeled data domain.In this paper,we propose an unsupervised person Re-ID method based on multi-granularity feature representation and domain adaptive learning,which can effectively im-prove the performance of unsupervised person re-identi-fication.The multi-granularity feature extraction module integrates global and local information of different granu-larity to obtain the multi-granularity person feature rep-resentation with rich discriminative information.The source domain classification module learns the labeled source dataset classification and obtains the person's dis-criminative knowledge in the source domain.On this basis,the domain adaptive module further considers the difference between the target domain and the source do-main to learn adaptively for the model.Experiments on multiple public datasets show that the proposed method can achieve a competitive performance among other state-of-the-art unsupervised Re-ID methods.
文献关键词：
中图分类号：
作者姓名：
FU Lihua;DU Yubin;DING Yu;WANG Dan;JIANG Hanxu;ZHANG Haitao
作者机构：
Beijing University of Technology,Beijing 100124,China
文献出处：
引用格式：
[1]FU Lihua;DU Yubin;DING Yu;WANG Dan;JIANG Hanxu;ZHANG Haitao-.Domain Adaptive Learning with Multi-Granularity Features for Unsupervised Person Re-identification)[J].电子学报（英文）,2022(01):116-128
A类：
granu,criminative
B类：
Domain,Adaptive,Learning,Multi,Granularity,Features,Unsupervised,Person,Re,identification,person,ID,aims,improve,model,scalability,better,results,unlabeled,domain,In,this,paper,unsupervised,granularity,feature,representation,learning,which,can,effectively,performance,extraction,module,integrates,global,local,information,different,rich,discriminative,source,classification,learns,obtains,knowledge,On,basis,further,considers,difference,between,target,adaptively,Experiments,multiple,public,datasets,show,that,proposed,achieve,competitive,among,other,state,art,methods
AB值：
0.440157
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
