典型文献
Unsupervised person re-identification based on removal of camera bias and dynamic updating of the memory bank
文献摘要：
In recent years,unsupervised person reidentification technology has made great strides.The technology re-trieves images of interested persons under different cameras from massive repositories of unlabeled images.However,in the current research,there are some existing problems,such as the influence of pedestrians appearing across cameras and pseudo-label noise.To solve these problems,we conduct research in two ways:removing the camera bias and dynamic-ally updating the memory model.In removing the camera bias,based on a learnable channel attention module,the features that are only related to cameras can be extracted from the feature map,thereby removing the camera bias in the global fea-tures and obtaining the features that can represent the pedestrians.In regards to dynamically updating the memory model,since the instance features do not necessarily belong to the identity represented by the pseudo-label,we adopt a method to update the memory dynamically according to the distance between the instance features and the category features so that the category features tend to be true.We combine the removal of the camera bias and the dynamic updating of the memory model to better solve problems in this field.Extensive experimentation demonstrates the superiority of our method over the state-of-the-art approaches on fully unsupervised Re-ID tasks.
文献关键词：
中图分类号：
作者姓名：
Jun Zhang;Xinmei Tian
作者机构：
Department of Electronic Engineering and Information Science,University of Science and Technology of China,Hefei 230027,China
文献出处：
引用格式：
[1]Jun Zhang;Xinmei Tian-.Unsupervised person re-identification based on removal of camera bias and dynamic updating of the memory bank)[J].中国科学技术大学学报,2022(12):58-67
A类：
reidentification,trieves,repositories
B类：
Unsupervised,removal,bias,updating,memory,bank,In,recent,years,unsupervised,technology,has,made,great,strides,images,interested,persons,under,different,cameras,from,massive,unlabeled,However,current,research,are,some,existing,problems,such,influence,pedestrians,appearing,across,pseudo,noise,To,solve,these,conduct,two,ways,removing,model,learnable,channel,attention,module,features,that,only,related,can,extracted,map,thereby,global,obtaining,regards,dynamically,since,instance,not,necessarily,belong,identity,represented,adopt,method,update,according,distance,between,category,tend,true,We,combine,better,this,field,Extensive,experimentation,demonstrates,superiority,our,over,state,art,approaches,fully,Re,ID,tasks
AB值：
0.479926
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。
