Viewpoint-Aware Loss(RK)

Reported on 6 benchmarks across 1 task · 1 paper · 1 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-IdentificationonDukeMTMC-reID
Rank-5· uses extra data· 2019-12-03
96.5
SOTA
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300
Person Re-IdentificationonMarket-1501
Rank-1· 2019-12-03
96.79
best: 98 (st-ReID(RE, RK))
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300
Person Re-IdentificationonMarket-1501
Rank-5· 2019-12-03
98.31
best: 98.9 (st-ReID(RE, RK))
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300
Person Re-IdentificationonMarket-1501
mAP· 2019-12-03
95.43
best: 96.21 (Unsupervised Pre-training (ResNet101+RK))
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300
Person Re-IdentificationonDukeMTMC-reID
Rank-1· uses extra data· 2019-12-03
93.9
best: 95.6 (CTL Model (ResNet50, 256x128))
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300
Person Re-IdentificationonDukeMTMC-reID
mAP· uses extra data· 2019-12-03
91.8
best: 97.1 (DenseIL)
Viewpoint-Aware Loss with Angular Regularization for Person Re-Identification arXiv:1912.01300