Adaptive L2 Regularization (with re-ranking)

Reported on 6 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision6 results

Person Re-IdentificationonMSMT17
Rank-1· 2020-07-15
84.9
best: 91.7 (SOLIDER (with re-ranking))
SOTA
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875
Person Re-IdentificationonMSMT17
mAP· 2020-07-15
76.7
best: 86.7 (CLIP-ReID (with re-ranking))
SOTA
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875
Person Re-IdentificationonMarket-1501
Rank-1· 2020-07-15
96
best: 98 (st-ReID(RE, RK))
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875
Person Re-IdentificationonMarket-1501
mAP· 2020-07-15
94.4
best: 96.21 (Unsupervised Pre-training (ResNet101+RK))
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875
Person Re-IdentificationonDukeMTMC-reID
Rank-1· 2020-07-15
92.2
best: 95.6 (CTL Model (ResNet50, 256x128))
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875
Person Re-IdentificationonDukeMTMC-reID
mAP· 2020-07-15
90.7
best: 97.1 (DenseIL)
Adaptive L2 Regularization in Person Re-Identification arXiv:2007.07875