Person Re-identification on Market-1501 1.0 (test)
97.4Rank-1DynaMix
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DynaMixTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 97.4 | — | — | 93.8 | |
| ReTextTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 97.4 | — | — | 93.8 | |
| TransReID-SSLBackbone=ViT1-B, Pre-training=LUPerson, Resolution=384x1282021.11 | 96.7 | — | — | 93.2 | |
| MoCoV2Backbone=MGN (ResNet-50), Pre-training=LUPerson, Resolution=384x1282021.11 | 96.4 | — | — | 91 | |
| TransReID-SSLBackbone=ViT1-S, Pre-training=LUPerson, Resolution=384x1282021.11 | 96.3 | — | — | 91.7 | |
| TransReID-SSLBackbone=ViT1-S, Pre-training=LUPerson, Resolution=Standard2021.11 | 96.2 | — | — | 91.3 | |
| ReMixTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 96.2 | — | — | 89.8 | |
| Local CNN2020.01 | 95.9 | — | — | 87.4 | |
| MGNf-dim=2,048, trained with classification and triplet losses=true2019.10 | 95.7 | — | — | 86.9 | |
| MGN2020.01 | 95.7 | — | — | 86.9 | |
| Pyramid2020.01 | 95.7 | — | — | 88.2 | |
| SCSNBackbone=ResNet-50, Pre-training=ImageNet, Resolution=384x1282021.11 | 95.7 | — | — | 88.5 | |
| SONA2020.01 | 95.68 | — | — | 88.67 | |
| ABD2020.01 | 95.6 | — | — | 88.28 | |
| PLR-OSNetbackbone=OSNet, parameters=3.4M2020.01 | 95.6 | — | — | 88.9 | |
| BOTBackbone=ResNet50-IBN-a, Pre-training=LUPerson, Resolution=Standard2021.11 | 95.6 | — | — | 90.1 | |
| ABDNetBackbone=ResNet-50, Pre-training=ImageNet, Resolution=384x1282021.11 | 95.6 | — | — | 88.3 | |
| BDB2020.01 | 95.3 | — | — | 86.7 | |
| P2-Netloss_function=triplet loss2019.10 | 95.2 | 98.2 | 99.1 | 85.6 | |
| IS-GANf-dim=2,0482019.10 | 95.2 | — | — | 87.1 | |
| TransReIDBackbone=ViT-B, Pre-training=ImageNet-21K, Resolution=384x1282021.11 | 95.2 | — | — | 89.5 | |
| MHN2020.01 | 95.1 | — | — | 85 | |
| MGNBackbone=ResNet-50, Pre-training=ImageNet, Resolution=384x1282021.11 | 95.1 | — | — | 87.5 | |
| BOTBackbone=ResNet50-IBN-a, Pre-training=ImageNet, Resolution=Standard2021.11 | 95 | — | — | 88.2 | |
| DCACTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 94.9 | — | — | 86.8 | |
| DG-Netf-dim=1,0242019.10 | 94.8 | — | — | 86 | |
| OSNetparameters=3.4M2020.01 | 94.8 | — | — | 84.9 | |
| DG-NetGenerated data usage=Joint learning2019.04 | 94.8 | — | — | 86 | |
| CAMA2020.01 | 94.7 | — | — | 84.5 | |
| MGNf-dim=2,048, trained with classification and triplet losses=true, our implementation=true2019.10 | 94.5 | — | — | 84.8 | |
| BagOfTricks2020.01 | 94.5 | — | — | 85.9 | |
| Auto-ReID2020.01 | 94.5 | — | — | 85.1 | |
| IAN2020.01 | 94.4 | — | — | 83.1 | |
| HPMf-dim=3,8402019.10 | 94.2 | — | — | 82.7 | |
| P2-Net2019.10 | 94 | 98 | 98.8 | 83.4 | |
| ReTextTraining Dataset=MSMT17-merged, Protocol=Protocol 12026.02 | 93.9 | — | — | 84.7 | |
| PCB + RPP2019.10 | 93.8 | 97.5 | 98.5 | 81.6 | |
| PCB+RPPf-dim=12,2882019.10 | 93.8 | — | — | 81.6 | |
| PCB2020.01 | 93.8 | — | — | 81.6 | |
| PCBGenerated data usage=None2019.04 | 93.8 | — | — | 81.6 | |
| ReTextTraining Dataset=MSMT17, Protocol=Protocol 12026.02 | 93.6 | — | — | 83.6 | |
| CRF2020.01 | 93.5 | — | — | 81.6 | |
| HDLF2020.01 | 93.3 | — | — | 79.1 | |
| Mancs2020.01 | 93.1 | — | — | 82.3 | |
| MancsGenerated data usage=None2019.04 | 93.1 | — | — | 82.3 | |
| GSRW2019.10 | 92.7 | 96.9 | 98.1 | 82.5 | |
| DynaMixTraining Dataset=MSMT17-merged, Protocol=Protocol 12026.02 | 92.6 | — | — | 80.8 | |
| SPREID2019.10 | 92.5 | 97.2 | 98.1 | 81.3 | |
| PATTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 92.4 | — | — | 81.5 | |
| SGGNN2019.10 | 92.3 | 96.1 | 97.4 | 82.8 | |
| PCBf-dim=12,2882019.10 | 92.3 | — | — | 77.4 | |
| Part-Alignedf-dim=5122019.10 | 91.7 | — | — | 79.6 | |
| Part-alignedGenerated data usage=None2019.04 | 91.7 | — | — | 79.6 | |
| ReTextTraining Dataset=RandPerson, Protocol=Protocol 12026.02 | 91.7 | — | — | 79.7 | |
| QAConv-GSTraining Dataset=Market-1501, Protocol=Protocol 12026.02 | 91.6 | — | — | 75.5 | |
| HA-CNN2019.10 | 91.2 | — | — | 75.7 | |
| HA-CNNf-dim=1,0242019.10 | 91.2 | — | — | 75.7 | |
| HA-CNN2020.01 | 91.2 | — | — | 75.7 | |
| HA-CNNGenerated data usage=None2019.04 | 91.2 | — | — | 75.7 | |
| DynaMixTraining Dataset=MSMT17, Protocol=Protocol 12026.02 | 91.1 | — | — | 77.7 | |
| FD-GANf-dim=2,0482019.10 | 90.5 | — | — | 77.7 | |
| FD-GANGenerated data usage=Separately generated images2019.04 | 90.5 | — | — | 77.7 | |
| PABR2019.10 | 90.2 | 96.1 | 97.4 | 76 | |
| KPM2020.01 | 90.1 | — | — | 75.3 | |
| MLFNf-dim=1,0242019.10 | 90 | — | — | 74.3 | |
| MLFN2020.01 | 90 | — | — | 74.4 | |
| MLFNGenerated data usage=None2019.04 | 90 | — | — | 74.3 | |
| GLADGenerated data usage=None2019.04 | 89.9 | — | — | 73.9 | |
| PN-GANf-dim=1,0242019.10 | 89.4 | — | — | 72.6 | |
| PN-GANGenerated data usage=Separately generated images2019.04 | 89.4 | — | — | 72.6 | |
| DynaMixTraining Dataset=RandPerson, Protocol=Protocol 12026.02 | 89.3 | — | — | 72.4 | |
| DPFL2019.10 | 88.9 | — | — | 73.1 | |
| SNL2020.01 | 88.27 | — | — | 73.43 | |
| PSE2019.10 | 87.7 | 94.5 | 96.8 | 69 | |
| PTGenerated data usage=Separately generated images2019.04 | 87.7 | — | — | 68.9 | |
| APRBackbone=Res502017.03 | 87.04 | 95.1 | 96.42 | 66.89 | |
| DaRef-dim=1282019.10 | 86.4 | — | — | 69.3 | |
| AACNPublish=CVPR2018, Backbone=GoogLeNet2017.03 | 85.9 | — | — | 66.87 | |
| AACN2019.10 | 85.9 | — | — | 66.9 | |
| Multi-pseudoGenerated data usage=Separately generated images2019.04 | 85.8 | — | — | 67.5 | |
| APR (w/o ARM)Backbone=Res50, attribute_re-weighting_module=false2017.03 | 85.71 | 94.32 | 96.46 | 66.59 | |
| JLML2019.10 | 85.1 | — | — | 65.5 | |
| PDC2019.10 | 84.1 | 92.7 | 94.9 | 63.4 | |
| LSROGenerated data usage=Separately generated images2019.04 | 84 | — | — | 66.1 | |
| ReMixTraining Dataset=MSMT17-merged, Protocol=Protocol 12026.02 | 84 | — | — | 61 | |
| GANPublish=ICCV2017, Backbone=Res502017.03 | 83.97 | — | — | 66.07 | |
| MGCAM2019.10 | 83.8 | — | — | 74.3 | |
| PANGenerated data usage=None2019.04 | 82.8 | — | — | 63.4 | |
| DSRPublish=CVPR2018, Backbone=Res502017.03 | 82.72 | — | — | 61.25 | |
| TransMatcherTraining Dataset=MSMT17-merged, Protocol=Protocol 12026.02 | 82.6 | — | — | 58.4 | |
| QAConv-GSTraining Dataset=MSMT17-merged, Protocol=Protocol 12026.02 | 82.4 | — | — | 56.9 | |
| SVDNetPublish=ICCV2017, Backbone=Res502017.03 | 82.3 | 92.3 | 95.2 | 62.1 | |
| SVDNet2019.10 | 82.3 | 92.3 | 95.2 | 62.1 | |
| DaF2019.10 | 82.3 | — | — | 72.4 | |
| SVDNetf-dim=2,0482019.10 | 82.3 | — | — | 62.1 | |
| SVDNetGenerated data usage=None2019.04 | 82.3 | — | — | 62.1 | |
| SSMGenerated data usage=None2019.04 | 82.2 | — | — | 68.8 | |
| EBBPublish=CVPR2018, Backbone=Inception2017.03 | 81.2 | 94.6 | 97 | — | |
| APR (w/o attri)Backbone=Res50, attribute_recognition_loss=false2017.03 | 81.03 | 91.29 | 94.28 | 58.74 | |
| Part-alignPublish=ICCV2017, Backbone=GoogLeNet2017.03 | 81 | 92 | 94.7 | 63.4 |