Joint Discriminative and Generative Learning for Person Re-identification

Benchmark Model Rank Results
person-re-identification-on-cuhk03DG-Net#9MAP: 61.1Rank-1: 65.6
person-re-identification-on-dukemtmc-reidDG-Net(RK)#16mAP: 88.31Rank-1: 90.26
person-re-identification-on-dukemtmc-reidDG-Net#53mAP: 74.8Rank-1: 86.6
person-re-identification-on-market-1501DG-Net(RK)#51Rank-1: 95.4mAP: 92.49
person-re-identification-on-market-1501DG-Net#65Rank-1: 94.8mAP: 86.0
person-re-identification-on-market-1501-cDG-Net#11Rank-1: 31.75mAP: 9.96mINP: 0.35
person-re-identification-on-msmt17DG-Net#30mAP: 52.3Rank-1: 77.2Rank-10: 90.5Rank-5: 87.4
unsupervised-domain-adaptation-on-duke-toDG-Net#15mAP: 26.83rank-1: 56.12rank-5: 72.18rank-10: 78.12
unsupervised-domain-adaptation-on-duke-to-1DG-Net#9mAP: 6.35rank-1: 20.59rank-5: 31.67rank-10: 37.04
unsupervised-domain-adaptation-on-market-toDG-Net#15mAP: 24.25rank-1: 42.62rank-5: 58.57rank-10: 64.63
unsupervised-domain-adaptation-on-market-to-1DG-Net#12mAP: 5.41rank-1: 17.11rank-5: 26.66rank-10: 31.62
unsupervised-person-re-identification-onDGNet#5mAP: 24.25Rank-1: 42.62Rank-5: 58.57Rank-10: 64.63
unsupervised-person-re-identification-on-1DGNet#5Rank-1: 56.12Rank-10: 72.18Rank-5: 78.12mAP: 26.83
unsupervised-person-re-identification-on-2DG-Net#6mAP: 5.41Rank-1: 17.11Rank-10: 26.66Rank-5: 31.62
unsupervised-person-re-identification-on-3DGNet#6mAP: 6.35Rank-1: 20.59Rank-10: 31.67Rank-5: 37.04