Human-Object Interaction Detection on V-COCO (test)
72.1AP (Role, Scenario 1)RLIPv2
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| RLIPv2Backbone=Swin-L2024.08 | 72.1 | — | — | — | — | 74.1 | — | — | |
| STIPMethod Type=Two-stage, Backbone=ResNet-502023.04 | 66 | — | — | — | — | 70.7 | — | — | |
| STIPBackbone=ResNet-502024.08 | 66 | — | — | — | — | 70.7 | — | — | |
| RLIPv2Method Type=One-stage, Supervision Setting=Fully-supervised2024.10 | 65.9 | — | — | — | — | 68 | — | — | |
| AGERMethod Type=One-stage, Supervision Setting=Fully-supervised2024.10 | 65.7 | — | — | — | — | 69.7 | — | — | |
| OCNCategory=One-stage, Backbone=R1012022.02 | 65.3 | — | — | — | — | 67.1 | — | — | |
| FleVRSBackbone=Focal-L2024.08 | 65.2 | — | — | — | — | 66.5 | — | — | |
| UniVRDBackbone=ViT-L2024.08 | 65.1 | — | — | — | — | 66.3 | — | — | |
| SSRTMethod Type=One-stage, Backbone=ResNet-1012023.04 | 65 | — | — | — | — | 67.1 | — | — | |
| LogicHOIMethod Type=One-stage, Supervision Setting=Fully-supervised2024.10 | 64.4 | — | — | — | — | 65.6 | — | — | |
| OCNCategory=One-stage, Backbone=R502022.02 | 64.2 | — | — | — | — | 66.3 | — | — | |
| ERNetBackbone=EfficientNetV2-XL2024.08 | 64.2 | — | — | — | — | — | — | — | |
| PViC w/ H-DETRBackbone=Swin-L, Detector=H-DETR2023.08 | 64.1 | — | — | — | — | 70.2 | — | — | |
| PVICBackbone=Swin-L2024.08 | 64.1 | — | — | — | — | 70.2 | — | — | |
| CDN-LArchitecture=One-stage, Extra features=None2021.08 | 63.91 | — | — | — | — | 65.89 | — | — | |
| CDNMethod Type=One-stage, Backbone=ResNet-1012023.04 | 63.9 | — | — | — | — | 65.9 | — | — | |
| CDNBackbone=ResNet-101, Detector=DETR2023.08 | 63.9 | — | — | — | — | 65.9 | — | — | |
| CDNBackbone=ResNet-1012024.08 | 63.9 | — | — | — | — | 65.8 | — | — | |
| GEN-VLKTMethod Type=One-stage, Backbone=ResNet-1012023.04 | 63.6 | — | — | — | — | 65.9 | — | — | |
| GEN-VLKTBackbone=ResNet-1012024.08 | 63.6 | — | — | — | — | 65.9 | — | — | |
| GEN-VLKT_lAnchor=Query, Scale=Large2022.03 | 63.58 | — | — | — | — | 65.93 | — | — | |
| DOQMethod Type=One-stage, Backbone=ResNet-502023.04 | 63.5 | — | — | — | — | — | — | — | |
| HOICLIPMethod Type=One-stage, Supervision Setting=Fully-supervised2024.10 | 63.5 | — | — | — | — | 64.8 | — | — | |
| GEN-VLKT_mAnchor=Query, Scale=Medium2022.03 | 63.28 | — | — | — | — | 65.58 | — | — | |
| QPIC + CPCBackbone=R502022.04 | 63.1 | — | — | — | — | 65.4 | — | — | |
| IFMethod Type=One-stage, Backbone=ResNet-502023.04 | 63 | — | — | — | — | 65.2 | — | — | |
| ParMapMethod Type=One-stage, Backbone=ResNet-502023.04 | 63 | — | — | — | — | 65.1 | — | — | |
| PViC w/ DETRBackbone=ResNet-50, Detector=DETR2023.08 | 62.8 | — | — | — | — | 67.8 | — | — | |
| GEN-VLKT_sAnchor=Query, Scale=Small2022.03 | 62.41 | — | — | — | — | 64.46 | — | — | |
| GEN-VLKTBackbone=ResNet-50, Detector=DETR2023.08 | 62.4 | — | — | — | — | 64.5 | — | — | |
| GEN-VLKTMethod Type=One-stage, Supervision Setting=Fully-supervised2024.10 | 62.4 | — | — | — | — | 64.5 | — | — | |
| CDN-BArchitecture=One-stage, Extra features=None2021.08 | 62.29 | — | — | — | — | 64.42 | — | — | |
| ViPLO_LMethod Type=Two-stage, Backbone=ViT-B/162023.04 | 62.2 | — | — | — | — | 68 | — | — | |
| MSTRMethod Type=One-stage, Backbone=ResNet-502023.04 | 62 | — | — | — | — | 65.2 | — | — | |
| MSTRBackbone=ResNet-50, Detector=DETR2023.08 | 62 | — | — | — | — | 65.2 | — | — | |
| RLIPBackbone=ResNet-50, Detector=DETR2023.08 | 61.9 | — | — | — | — | 64.2 | — | — | |
| RLIPBackbone=ResNet-502024.08 | 61.9 | — | — | — | — | 64.2 | — | — | |
| CDN-SArchitecture=One-stage, Extra features=None2021.08 | 61.68 | — | — | — | — | 63.77 | — | — | |
| HOTR + CPCBackbone=R502022.04 | 61.6 | — | — | — | — | 66.8 | — | — | |
| UPTBackbone=ResNet-101-DC52021.12 | 61.3 | — | — | — | — | 67.1 | — | — | |
| UPTMethod Type=Two-stage, Backbone=ResNet-1012023.04 | 61.3 | — | — | — | — | 67.1 | — | — | |
| UPTBackbone=ResNet-101-DC5, Detector=DETR2023.08 | 61.3 | — | — | — | — | 67.1 | — | — | |
| ViPLO_SMethod Type=Two-stage, Backbone=ViT-B/322023.04 | 60.9 | — | — | — | — | 66.6 | — | — | |
| UPTBackbone=ResNet-1012021.12 | 60.7 | — | — | — | — | 66.2 | — | — | |
| FGAHOIArchitecture Type=One-stage Method, Backbone=Swin-Tiny2023.01 | 60.5 | — | — | — | — | 61.2 | — | — | |
| EZ-HOIMethod Type=Two-stage, Supervision Setting=Fully-supervised2024.10 | 60.5 | — | — | — | — | 66.2 | — | — | |
| UPTBackbone=ResNet-502021.12 | 59 | — | — | — | — | 64.5 | — | — | |
| FGAHOIArchitecture Type=One-stage Method, Backbone=ResNet-502023.01 | 59 | — | — | — | — | 59.3 | — | — | |
| UPTMethod Type=Two-stage, Supervision Setting=Fully-supervised2024.10 | 59 | — | — | — | — | 64.5 | — | — | |
| QPICBackbone=ResNet-502021.03 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICBackbone=ResNet-502021.03 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICArchitecture=One-stage, Extra features=None2021.08 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICCategory=One-stage, Backbone=R502022.02 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICBackbone=ResNet-1012021.12 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICBackbone=R502022.04 | 58.8 | — | — | — | — | 61 | — | — | |
| QPIC-R50Anchor=Query, Backbone=ResNet-502022.03 | 58.8 | — | — | — | — | 61 | — | — | |
| QPICBackbone=ResNet-101, Detector=DETR2023.08 | 58.8 | — | — | — | — | 61 | — | — | |
| ADA-CMMethod Type=Two-stage, Supervision Setting=Fully-supervised2024.10 | 58.6 | — | — | — | — | 64 | — | — | |
| QPICBackbone=ResNet-1012021.03 | 58.3 | — | — | — | — | 60.7 | — | — | |
| QPICBackbone=ResNet-1012021.03 | 58.3 | — | — | — | — | 60.7 | — | — | |
| QPIC-R101Anchor=Query, Backbone=ResNet-1012022.03 | 58.3 | — | — | — | — | 60.7 | — | — | |
| QPICMethod Type=One-stage, Backbone=ResNet-1012023.04 | 58.3 | — | — | — | — | 60.7 | — | — | |
| QAHOIArchitecture Type=One-stage Method, Backbone=ResNet-502023.01 | 58.2 | — | — | — | — | 58.7 | — | — | |
| DIRVArchitecture=One-stage, Extra features=None2021.08 | 56.1 | — | — | — | — | — | — | — | |
| DIRVCategory=One-stage, Backbone=ED-d32022.02 | 56.1 | — | — | — | — | — | — | — | |
| DIRVArchitecture Type=One-stage Method2023.01 | 56.1 | — | — | — | — | — | — | — | |
| DIRVBackbone=EfficientDet-d32024.08 | 56.1 | — | — | — | — | — | — | — | |
| HOTRBackbone=R502021.04 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRArchitecture=One-stage, Extra features=None2021.08 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRCategory=One-stage, Backbone=R502022.02 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRBackbone=ResNet-502021.12 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRBackbone=R502022.04 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRAnchor=Query2022.03 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTR (transformer method)Backbone=ResNet-502022.07 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRBackbone=ResNet-502022.07 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRArchitecture Type=One-stage Method2023.01 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRMethod Type=One-stage, Backbone=ResNet-502023.04 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRBackbone=ResNet-50, Detector=DETR2023.08 | 55.2 | — | — | — | — | 64.4 | — | — | |
| HOTRBackbone=ResNet-502024.08 | 55.2 | — | — | — | — | 64.4 | — | — | |
| GG-NetArchitecture=One-stage, Extra features=None2021.08 | 54.7 | — | — | — | — | — | — | — | |
| GGNetCategory=One-stage, Backbone=HOG1042022.02 | 54.7 | — | — | — | — | — | — | — | |
| GG-NetAnchor=Point2022.03 | 54.7 | — | — | — | — | — | — | — | |
| GG-NetMethod Type=One-stage, Backbone=Hourglass-1042023.04 | 54.7 | — | — | — | — | — | — | — | |
| SCGBackbone=ResNet-50-FPN2021.12 | 54.2 | — | — | — | — | 60.9 | — | — | |
| SCGMethod Type=Two-stage, Backbone=ResNet-50-FPN2023.04 | 54.2 | — | — | — | — | 60.9 | — | — | |
| SCGBackbone=ResNet-50-FPN, Detector=DETR2023.08 | 54.2 | — | — | — | — | 60.9 | — | — | |
| AS-NetArchitecture=One-stage, Extra features=None2021.08 | 53.9 | — | — | — | — | — | — | — | |
| ASNetCategory=One-stage, Backbone=R502022.02 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetBackbone=ResNet-502021.12 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetBackbone=R502022.04 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetAnchor=Query2022.03 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetArchitecture Type=One-stage Method2023.01 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetMethod Type=One-stage, Backbone=ResNet-502023.04 | 53.9 | — | — | — | — | — | — | — | |
| AS-NetBackbone=ResNet-50, Detector=DETR2023.08 | 53.9 | — | — | — | — | — | — | — | |
| IDN2020.10 | 53.3 | — | — | — | — | 60.3 | — | — | |
| IDNBackbone=R502021.04 | 53.3 | — | — | — | — | 60.3 | — | — | |
| IDNArchitecture=Two-stage, Extra features=None2021.08 | 53.3 | — | — | — | — | 60.3 | — | — | |
| IDNCategory=Two-stage, Backbone=R502022.02 | 53.3 | — | — | — | — | 60.3 | — | — | |
| IDNBackbone=ResNet-502021.12 | 53.3 | — | — | — | — | 60.3 | — | — | |
| PDNetBackbone=R50-FPN2022.04 | 53.3 | — | — | — | — | — | — | — |