Vehicle Re-identification on VeRi-776 (test)
97.7Rank-1ProNet++
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ProNet++Input Resolution=256 x 256, Extra data or annotations=False2023.08 | 97.7 | 83.4 | — | — | |
| VOC ReID2021.04 | 97.6 | 82.8 | — | — | |
| CLIP-ReIDLearning paradigm=Fully supervised2023.10 | 97.4 | 83.3 | — | — | |
| CLIP-ReIDBackbone=ViT2022.11 | 97.4 | 83.3 | — | — | |
| CLIP-ReIDBackbone=ViT, Side Information Embedding (SIE)=true, Overlapping Layer Patch (OLP)=true2022.11 | 97.3 | 84.5 | — | — | |
| TransReIDLearning paradigm=Fully supervised2023.10 | 97.1 | 82 | — | — | |
| PCL-CLIP Lpel+LidLearning paradigm=Fully supervised, Loss=Lpel + Lid2023.10 | 97.1 | 82.5 | 98.6 | 99.2 | |
| TransReIDBackbone=ViT, Side Information Embedding (SIE)=true, Overlapping Layer Patch (OLP)=true2022.11 | 97.1 | 82 | — | — | |
| MALWbackbone=ResNeXt101_ibn_a, multi-head=true, losses=CE, SupCon, MALW, synthetic data=false, post-processing=false2021.04 | 97 | 87.1 | — | — | |
| PCL-CLIP LpelLearning paradigm=Fully supervised, Loss=Lpel2023.10 | 97 | 81.1 | 98.7 | 99.3 | |
| TransReIDBackbone=ViT-Base, Side Information=false, Input size=256x2562022.05 | 96.9 | 79.2 | — | — | |
| ViT-Base + DCALBackbone=ViT-Base, Input size=256x2562022.05 | 96.9 | 80.2 | — | — | |
| DCALBackbone=ViT-B/16 (Transformer), Extra data or annotations=True, Input Resolution=256 x 2562023.08 | 96.9 | 80.2 | — | — | |
| TransReIDBackbone=ViT2022.11 | 96.9 | 80.6 | — | — | |
| DCALBackbone=ViT2022.11 | 96.9 | 80.2 | — | — | |
| GiTCategory=Proposed2021.07 | 96.86 | 80.34 | — | — | |
| MSINetParams=2.3M, Inference Time=0.71x2023.03 | 96.8 | 78.8 | — | — | |
| TransReIDBackbone=ViT-B/16 (Transformer), Extra data or annotations=True, Input Resolution=256 x 2562023.08 | 96.8 | 81.2 | — | — | |
| ProNetInput Resolution=256 x 256, Extra data or annotations=False2023.08 | 96.8 | 81 | — | — | |
| VehicleNetLearning paradigm=Fully supervised2023.10 | 96.8 | 83.4 | — | — | |
| TiF learner w/o cBackbone=OpenCLIP, # Shots=162024.03 | 96.8 | — | — | — | |
| CLIP-ReIDBackbone=CNN2022.11 | 96.8 | 80.3 | — | — | |
| HPGNReferences=Proposed2020.05 | 96.72 | 80.18 | — | — | |
| GB+GFB+SLBExtra data or annotations=False, Input Resolution=256 x 2562023.08 | 96.7 | 81 | — | — | |
| SCA2020.12 | 96.54 | 87.48 | — | — | |
| DeiT-Base + DCALBackbone=DeiT-Base, Input size=256x2562022.05 | 96.5 | 80 | — | — | |
| ViT-BParams=~86M, Inference Time=1.79x2023.03 | 96.5 | 78.2 | — | — | |
| PGANExtra data or annotations=False2023.08 | 96.5 | 79.3 | — | — | |
| PGANBackbone=CNN2022.11 | 96.5 | 79.3 | — | — | |
| GNN-based re-ranking2020.12 | 96.42 | 88.61 | — | — | |
| SAVERCategory=Global + Local (Pure CNN), Same training tricks as GiT=true2021.07 | 96.4 | 79.6 | — | — | |
| SAVERExtra data or annotations=False, Input Resolution=256 x 2562023.08 | 96.4 | 79.6 | — | — | |
| UFDNExtra data or annotations=False2023.08 | 96.4 | 81.5 | — | — | |
| SAVERBackbone=CNN2022.11 | 96.4 | 79.6 | — | — | |
| k-reciprocal2020.12 | 96.36 | 88.44 | — | — | |
| VOC-ReIDCategory=Global + Local (Pure CNN), Same training tricks as GiT=true2021.07 | 96.3 | 79.7 | — | — | |
| PROVID-BOT2020.07 | 96.1 | 77.2 | 97.9 | — | |
| ViT-BaseBackbone=ViT-Base, Input size=256x2562022.05 | 96 | 78.1 | — | — | |
| DeiT-Small + DCALBackbone=DeiT-Small, Input size=256x2562022.05 | 95.9 | 78.1 | — | — | |
| DeiT-BaseBackbone=DeiT-Base, Input size=256x2562022.05 | 95.9 | 78.3 | — | — | |
| DeiT-BParams=~86M, Inference Time=1.79x2023.03 | 95.9 | 78.4 | — | — | |
| ViTCategory=Transformer, Same training tricks as GiT=true2021.07 | 95.84 | 78.92 | — | — | |
| UMTSExtra data or annotations=False2023.08 | 95.8 | 75.9 | — | — | |
| UMTSBackbone=CNN2022.11 | 95.8 | 75.9 | — | — | |
| ResVKD-50bamBackbone=ResNet-50, Distillation=VKD, Attention=BAM2020.07 | 95.7 | 81.6 | 98 | — | |
| FIDIInput size=256x2562022.05 | 95.7 | 77.6 | — | — | |
| EIA-NetBackbone=CNN2022.11 | 95.7 | 79.3 | — | — | |
| FIDIBackbone=CNN2022.11 | 95.7 | 77.6 | — | — | |
| baselineBackbone=CNN2022.11 | 95.7 | 79.3 | — | — | |
| baselineBackbone=ViT2022.11 | 95.7 | 79.3 | — | — | |
| PVENInput size=256x2562022.05 | 95.6 | 79.5 | — | — | |
| PVENCategory=Global + Local (Pure CNN), Same training tricks as GiT=true2021.07 | 95.6 | 79.5 | — | — | |
| PVENExtra data or annotations=True, Input Resolution=256 x 2562023.08 | 95.6 | 79.5 | — | — | |
| PVENBackbone=CNN2022.11 | 95.6 | 79.5 | — | — | |
| baseline2020.12 | 95.59 | 78.94 | — | — | |
| DeiT-SmallBackbone=DeiT-Small, Input size=256x2562022.05 | 95.5 | 76.7 | — | — | |
| DeiT-SParams=~22M, Inference Time=0.97x2023.03 | 95.5 | 76.3 | — | — | |
| Appearance+LicenseReferences=ICIP 20192020.05 | 95.41 | 78.08 | — | — | |
| App+LicenseCategory=Global + Local (Pure CNN)2021.07 | 95.41 | 78.08 | — | — | |
| CALBackbone=ResNet-50, Input size=256x2562022.05 | 95.4 | 74.3 | — | — | |
| PCRNetCategory=Global + Local (CNN & GN), Same training tricks as GiT=true2021.07 | 95.4 | 78.6 | — | — | |
| CALExtra data or annotations=False, Input Resolution=256 x 2562023.08 | 95.4 | 74.3 | — | — | |
| CALLearning paradigm=Fully supervised2023.10 | 95.4 | 74.3 | 97.9 | — | |
| CALBackbone=CNN2022.11 | 95.4 | 74.3 | — | — | |
| SFF+SAttCategory=Global + Local (Pure CNN), Same training tricks as GiT=true2021.07 | 95.35 | 77.28 | — | — | |
| CFVMNetCategory=Global + Local (Pure CNN)2021.07 | 95.3 | 77.06 | — | — | |
| CFVMNetBackbone=CNN2022.11 | 95.3 | 77.1 | — | — | |
| ResVKD-50Backbone=ResNet-50, Distillation=VKD2020.07 | 95.2 | 82.2 | 98 | — | |
| SFF+SAttReferences=IJCNN 20192020.05 | 94.93 | 74.11 | — | — | |
| SANCategory=Global + Local (Pure CNN), Same training tricks as GiT=true2021.07 | 94.82 | 74.68 | — | — | |
| DeiT-Tiny + DCALBackbone=DeiT-Tiny, Input size=256x2562022.05 | 94.7 | 74.1 | — | — | |
| SGATCategory=Global + Local (CNN & GN), Same training tricks as GiT=true2021.07 | 94.65 | 76.32 | — | — | |
| Part RegularizationReferences=CVPR 20192020.05 | 94.3 | 74.3 | — | — | |
| DeiT-TinyBackbone=DeiT-Tiny, Input size=256x2562022.05 | 94.3 | 71.3 | — | — | |
| Part RegularCategory=Global + Local (Pure CNN)2021.07 | 94.3 | 74.3 | — | — | |
| PRNBackbone=CNN2022.11 | 94.3 | 74.3 | — | — | |
| SPANInput size=256x2562022.05 | 94 | 68.9 | — | — | |
| SPANBackbone=Transformer, Extra data or annotations=True2023.08 | 94 | 68.9 | — | — | |
| SPANBackbone=CNN2022.11 | 94 | 68.9 | — | — | |
| SAN2020.07 | 93.3 | 72.5 | 97.1 | — | |
| SANReferences=arXiv 20192020.05 | 93.3 | 72.5 | — | — | |
| SANInput size=256x2562022.05 | 93.3 | 72.5 | — | — | |
| SANLearning paradigm=Fully supervised2023.10 | 93.3 | 72.5 | — | — | |
| SANBackbone=CNN2022.11 | 93.3 | 72.5 | — | — | |
| PAMTRI2020.07 | 92.9 | 71.9 | 92.9 | — | |
| PAMTRIReferences=ICCV 20192020.05 | 92.86 | 71.88 | — | — | |
| PAMTRICategory=Global + Local (Pure CNN)2021.07 | 92.86 | 71.88 | — | — | |
| PAMTRI(ALL)2021.04 | 92.8 | 71.8 | — | — | |
| MLFN+TripletReferences=CVPRW 20192020.05 | 92.55 | 71.78 | — | — | |
| SFSCQuery Model=phi, Gallery Model=phi2022.06 | 92.32 | 66.55 | — | — | |
| Unified ModelQuery Model=phi, Gallery Model=phi2022.06 | 92.2 | 66.5 | — | — | |
| MTML+OSG+Re-rankingReferences=CVPRW 20192020.05 | 92 | 68.3 | — | — | |
| MRMReferences=Neurocomputing 20192020.05 | 91.77 | 68.55 | — | — | |
| MRMCategory=Global + Local (Pure CNN)2021.07 | 91.77 | 68.55 | — | — | |
| DMMLReferences=ICCV 20192020.05 | 91.2 | 70.1 | — | — | |
| DMMLCategory=Global + Local (Pure CNN)2021.07 | 91.2 | 70.1 | — | — | |
| TiF learner w/o cBackbone=OpenCLIP, # Shots=82024.03 | 91.2 | — | — | — | |
| PCL-CLIP O2CAPLearning paradigm=Unsupervised, Protocol=O2CAP2023.10 | 90.7 | 45.5 | 93.9 | 95 | |
| SFSCQuery Model=phi_9/16, Gallery Model=phi2022.06 | 90.48 | 62.72 | — | — | |
| BCT-SQuery Model=phi_9/16, Gallery Model=phi2022.06 | 90.42 | 58.2 | — | — |