Title
Deviation based clustering for unsupervised person re-identification.
Abstract
•Taking the inter- and intra-cluster deviation into consideration comprehensively.•Merging the closest clusters and grouping samples with the same identity.•Avoiding generating poor clusters containing samples from different identities.•Outperforming the state-of-the-art method BUC on Market-1501.
Year
DOI
Venue
2020
10.1016/j.patrec.2020.04.039
Pattern Recognition Letters
Keywords
DocType
Volume
Person re-identification,Neural networks,Clustering,Unsupervised learning
Journal
135
ISSN
Citations 
PageRank 
0167-8655
1
0.36
References 
Authors
0
4
Name
Order
Citations
PageRank
Munan Ning182.87
Zeng Kaiwei2101.56
Yang Guo36732.72
Yaohua Wang44414.23