本篇文章为目前专栏中总结论文的索引:
1.换衣方向
COCAS A Large-Scale Clothes Changing Person Dataset for Re-identification
Person Re-identification by Contour Sketch under Moderate Clothing Change
Fine-Grained Shape-Appearance Mutual Learning for Cloth-Changing Person Re-Identification
2.无监督(UDA)
Global Distance-distributions Separation for Unsupervised Person Re-identification
Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person
3.无监督域适应(DG)
Invariance Matters: Exemplar Memory for Domain AdaptivePerson Re-identification
Meta Batch-Instance Normalization for Generalizable Person Re-Identification
DomainMix: Learning Generalizable Person Re-Identification Without Human Annotations
4.步态识别方向
5.分块方向:
AlignedReID: Surpassing Human-Level Performance in Person Re-Identification
Relation Network for Person Re-identification
本文汇总了当前研究热点,包括大规模换衣场景下的人脸识别(COCASALarge)、服装变化下的重新识别方法,以及无监督和领域适应技术的最新进展,如全球距离分布分离、互惠教学伪标签精炼等。同时探讨了步态识别和分块方法的应用,以及基于GAN的方向。
902

被折叠的 条评论
为什么被折叠?



