Object Detection on OV-LVIS v1 (val)
37AP_mask_novelCoDet
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| CoDetBackbone=EVA02-L (304M), Supervision=Caption, Strict=true2023.10 | 37 | — | — | — | — | 46.3 | 46.3 | 44.7 | |
| F-VLMBackbone=R50x64 (420M), Supervision=CLIP, Strict=true2023.10 | 32.8 | — | — | — | — | — | — | 34.9 | |
| CoDetBackbone=SwinB (88M), Supervision=Caption, Strict=true2023.10 | 29.4 | — | — | — | — | 39.5 | 43 | 39.2 | |
| F-VLMBackbone=R50x4 (87M), Supervision=CLIP, Strict=true2023.10 | 26.3 | — | — | — | — | — | — | 28.5 | |
| VLDetBackbone=SwinB (88M), Supervision=Caption, Strict=true2023.10 | 26.3 | — | — | — | — | 39.4 | 41.9 | 38.1 | |
| OursBackbone Network=ViT-B/162026.04 | 25.5 | — | — | — | — | — | — | 24.7 | |
| CLIPSelfBackbone Network=ViT-B/162026.04 | 25.1 | — | — | — | — | — | — | 24.5 | |
| DeticBackbone=SwinB (88M), Supervision=Caption, Strict=false2023.10 | 23.9 | — | — | — | — | 40.2 | 42.8 | 38.4 | |
| CoDetBackbone=RN50, Supervision=Caption, Strict=true2023.10 | 23.4 | — | — | — | — | 30 | 34.6 | 30.7 | |
| BARONBackbone=RN50-FPN, Supervision=CLIP, Strict=true2023.10 | 22.6 | — | — | — | — | 27.6 | 29.8 | 27.6 | |
| RegionCLIPBackbone=R50x4 (87M), Supervision=Caption, Strict=true2023.10 | 22 | — | — | — | — | 32.1 | 36.9 | 32.3 | |
| VLDetBackbone=RN50, Supervision=Caption, Strict=true2023.10 | 21.7 | — | — | — | — | 29.8 | 34.3 | 30.1 | |
| OursBackbone Network=RN502026.04 | 21.7 | — | — | — | — | — | — | 26 | |
| SAS-DetBackbone Network=RN502026.04 | 20.9 | — | — | — | — | — | — | 26.1 | |
| OADPBackbone Network=RN502026.04 | 19.9 | — | — | — | — | — | — | — | |
| DetProBackbone=RN50-FPN, Supervision=CLIP, Strict=true2023.10 | 19.8 | — | — | — | — | 25.6 | 28.9 | 25.9 | |
| DetProBackbone Network=RN502026.04 | 19.8 | — | — | — | — | — | — | 25.9 | |
| DeticBackbone=RN50, Supervision=Caption, Strict=false2023.10 | 19.5 | — | — | — | — | — | — | 30.9 | |
| BARONBackbone Network=RN502026.04 | 19.2 | — | — | — | — | — | — | 26.5 | |
| PromptDetBackbone=RN50-FPN, Supervision=Caption, Strict=false2023.10 | 19 | — | — | — | — | 18.5 | 25.8 | 21.4 | |
| F-VLMBackbone=RN50-FPN, Supervision=CLIP, Strict=true2023.10 | 18.6 | — | — | — | — | — | — | 24.2 | |
| F-VLMBackbone Network=RN502026.04 | 18.6 | — | — | — | — | — | — | 24.2 | |
| OV-DETRBackbone=RN50-C4, Supervision=Caption, Strict=false2023.10 | 17.4 | — | — | — | — | 25 | 32.5 | 26.6 | |
| RegionCLIPBackbone=RN50-C4, Supervision=Caption, Strict=true2023.10 | 17.1 | — | — | — | — | 27.4 | 34 | 28.2 | |
| ViLDBackbone=RN50-FPN, Supervision=CLIP, Strict=true2023.10 | 16.6 | — | — | — | — | 24.6 | 30.3 | 25.5 | |
| ViLDBackbone Network=RN502026.04 | 16.6 | — | — | — | — | — | — | 25.5 | |
| BARONEnsemble=false, Learned Prompt=false2023.02 | — | 17.3 | 25.6 | 31 | 26.3 | — | — | — | |
| BARONEnsemble=true, Learned Prompt=false2023.02 | — | 20.1 | 28.4 | 32.2 | 28.4 | — | — | — | |
| BARONEnsemble=true, Learned Prompt=true2023.02 | — | 23.2 | 29.3 | 32.5 | 29.5 | — | — | — | |
| BARON2024.06 | — | 28.4 | 20.1 | 28.4 | 32.2 | — | — | — | |
| BARONLearnable prompt templates=true2024.06 | — | 29.5 | 23.2 | 29.3 | 32.5 | — | — | — | |
| DetProEnsemble=true, Learned Prompt=true2023.02 | — | 20.8 | 27.8 | 32.4 | 28.4 | — | — | — | |
| DetProLearnable prompt templates=true2024.06 | — | 28.4 | 20.8 | 27.8 | 32.4 | — | — | — | |
| LBP2024.06 | — | 29.1 | 22.2 | 28.8 | 32.4 | — | — | — | |
| LBPLearnable prompt templates=true2024.06 | — | 29.9 | 24.1 | 29.5 | 32.8 | — | — | — | |
| ViLDEnsemble=false, Learned Prompt=false2023.02 | — | 16.3 | 21.2 | 31.6 | 24.4 | — | — | — | |
| ViLDEnsemble=true, Learned Prompt=false2023.02 | — | 16.7 | 26.5 | 34.2 | 27.8 | — | — | — | |
| ViLD2024.06 | — | 27.8 | 16.7 | 26.5 | 34.2 | — | — | — | |
| ViLD*Ensemble=true, Learned Prompt=false2023.02 | — | 17.4 | 27.5 | 31.9 | 27.5 | — | — | — |