Visual Grounding on ReferCOCOg Google (val)
72.98Accuracy @ 0.5 IoUVLTVG
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VLTVG2023.03 | 72.98 | — | — | |
| QRNetStage=One-stage, Backbone=Swin-S2022.03 | 71.89 | — | — | |
| JointNLT2023.03 | 70.07 | — | — | |
| CM-Att-EraseStage=Two-stage, Backbone=ResNet-1012022.03 | 68.67 | — | — | |
| TransVGStage=One-stage, Backbone=Swin-S2022.03 | 67.81 | — | — | |
| TransVGStage=One-stage, Backbone=ResNet-1012022.03 | 67.02 | — | — | |
| TransVG2023.03 | 67.02 | — | — | |
| NMTreeStage=Two-stage, Backbone=ResNet-1012022.03 | 64.62 | — | — | |
| ReSC-LargeStage=One-stage, Backbone=DarkNet-532022.03 | 63.12 | — | — | |
| ReSC-Large2023.03 | 63.12 | — | — | |
| LBYL-NetStage=One-stage, Backbone=DarkNet-532022.03 | 62.7 | — | — | |
| LBYL-Net2023.03 | 62.7 | — | — | |
| VCStage=Two-stage, Backbone=VGG162022.03 | 62.3 | — | — | |
| LGRANSStage=Two-stage, Backbone=VGG162022.03 | 61.78 | — | — | |
| NMTree2023.03 | 61.78 | — | — | |
| FAOAStage=One-stage, Backbone=DarkNet-532022.03 | 56.12 | — | — | |
| APE (A)Backbone=ViT-L2023.12 | — | 39.4 | 28.4 | |
| APE (B)Backbone=ViT-L2023.12 | — | 80.5 | 64.3 | |
| APE (C)Backbone=ViT-L2023.12 | — | 79.5 | 62.6 | |
| APE (D)Backbone=ViT-L2023.12 | — | 79.9 | 63.3 | |
| JointNLT2024.03 | — | — | 70.07 | |
| LBYL2024.03 | — | — | 62.7 | |
| QueryNLT2024.03 | — | — | 72 | |
| ReSC2024.03 | — | — | 63.12 | |
| TransVG2024.03 | — | — | 67.02 | |
| VLTVG2024.03 | — | — | 73 |