Style Normalization and Restitution for Generalizable Person Re-identification

Existing fully-supervised person re-identification (ReID) methods usually suffer from poor generalization capability caused by domain gaps. The key to solving this problem lies in filtering out identity-irrelevant interference and learning domain-invariant person representations... (read more)

PDF Abstract CVPR 2020 PDF CVPR 2020 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Unsupervised Domain Adaptation Duke to Market SNR mAP 61.7 # 7
rank-1 82.8 # 8
rank-5 - # 14
rank-10 - # 16
Unsupervised Domain Adaptation Market to Duke SNR mAP 58.1 # 5
rank-1 76.3 # 5
rank-5 - # 15
rank-10 - # 14

Methods used in the Paper