Image Search on COCO (test)
7,310R@5m-CNNENS
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| m-CNNENSBackbone=VGG2015.04 | 7,310 | 4,280 | 8,410 | 2 | |
| m-RNN-vggBackbone=VGG2015.04 | 7,300 | 4,100 | 8,350 | 2 | |
| DVSA2015.04 | 6,990 | 3,840 | 8,050 | 1 | |
| m-CNNstBackbone=VGG2015.04 | 6,960 | 3,830 | 8,100 | 2 | |
| m-CNNENSBackbone=VGG2015.04 | 6,860 | 3,260 | 8,280 | 3 | |
| FV (GMM+HGLMM)2015.04 | 6,790 | 3,940 | 8,090 | 2 | |
| FV (LMM)2015.04 | 6,780 | 3,860 | 7,980 | 3 | |
| STV (combine-skip)2015.04 | 6,770 | 3,380 | 8,210 | 3 | |
| m-CNNphsBackbone=VGG2015.04 | 6,750 | 3,460 | 8,140 | 3 | |
| STV (bi-skip)2015.04 | 6,730 | 3,270 | 7,960 | 3 | |
| m-CNNphlBackbone=VGG2015.04 | 6,730 | 3,510 | 8,160 | 2 | |
| FV (GMM)2015.04 | 6,700 | 3,900 | 8,030 | 3 | |
| m-CNNwdBackbone=VGG2015.04 | 6,690 | 3,410 | 7,970 | 3 | |
| FV (HGLMM)2015.04 | 6,660 | 3,770 | 7,910 | 3 | |
| m-CNNwdBackbone=VGG2015.04 | 6,470 | 2,790 | 8,040 | 3 | |
| STV (uni-skip)2015.04 | 6,450 | 3,060 | 7,980 | 3 | |
| m-CNNphsBackbone=VGG2015.04 | 6,440 | 2,760 | 7,950 | 3 | |
| m-CNNstBackbone=VGG2015.04 | 6,340 | 2,740 | 7,950 | 3 | |
| m-CNNphlBackbone=VGG2015.04 | 6,280 | 2,710 | 7,930 | 3 | |
| FV (Mean Vec)2015.04 | 6,180 | 3,320 | 7,510 | 3 | |
| DVSA2015.04 | 6,020 | 2,740 | 7,480 | 3 | |
| STV (combine-skip)2015.04 | 6,000 | 2,590 | 7,460 | 4 | |
| FV (GMM+HGLMM)2015.04 | 5,980 | 2,510 | 7,660 | 4 | |
| FV (LMM)2015.04 | 5,950 | 2,500 | 7,610 | 4 | |
| FV (GMM)2015.04 | 5,920 | 2,420 | 7,600 | 4 | |
| FV (HGLMM)2015.04 | 5,880 | 2,490 | 7,650 | 4 | |
| STV (bi-skip)2015.04 | 5,710 | 2,420 | 7,320 | 4 | |
| STV (uni-skip)2015.04 | 5,640 | 2,270 | 7,170 | 4 | |
| FV (Mean Vec)2015.04 | 5,640 | 2,420 | 7,240 | 4 | |
| m-RNN-vggBackbone=VGG2015.04 | 4,220 | 2,900 | 7,700 | 3 | |
| VK-OOD2023.02 | 83.6 | — | — | — | |
| VinVL-BModel Scale=Base2023.02 | 83.2 | — | — | — | |
| ALBEF(4M)Pre-train images=4M2023.02 | 81.5 | — | — | — | |
| UNITER-BModel Scale=Base2023.02 | 78.5 | — | — | — | |
| PixelBERT2023.02 | 77.5 | — | — | — | |
| EVA-02-CLIP-E/14+zero-shot=true, backbone=ViT-E/14+2023.03 | 75 | 51.1 | 82.7 | — | |
| Open CLIP-G/14zero-shot=true, backbone=ViT-G/142023.03 | 74.9 | 51.4 | 83 | — | |
| EVA-02-CLIP-E/14zero-shot=true, backbone=ViT-E/142023.03 | 74.7 | 50.8 | 82.5 | — | |
| EVA-01-CLIP-g/14+zero-shot=true, backbone=ViT-g/14+2023.03 | 74 | 50.3 | 82.1 | — | |
| Open CLIP-H/14zero-shot=true, backbone=ViT-H/142023.03 | 73.4 | 49.5 | 81.5 | — | |
| Open CLIP-g/14zero-shot=true, backbone=ViT-g/142023.03 | 73.3 | 48.8 | 81.5 | — | |
| ViLT-BModel Scale=Base2023.02 | 72 | — | — | — | |
| EVA-02-CLIP-L/14+zero-shot=true, backbone=ViT-L/14+2023.03 | 71.7 | 47.9 | 80 | — | |
| EVA-02-CLIP-L/14zero-shot=true, backbone=ViT-L/142023.03 | 71.2 | 47.5 | 79.7 | — | |
| Open CLIP-L/14zero-shot=true, backbone=ViT-L/142023.03 | 70.7 | 46.1 | 79.4 | — | |
| EVA-01-CLIP-g/14zero-shot=true, backbone=ViT-g/142023.03 | 68.5 | 44.1 | 77.3 | — | |
| EVA-02-CLIP-B/16zero-shot=true, backbone=ViT-B/162023.03 | 66.9 | 42.2 | 76.3 | — | |
| Open CLIP-B/16zero-shot=true, backbone=ViT-B/162023.03 | 66.7 | 42.3 | 77.1 | — | |
| OpenAI CLIP-L/14+zero-shot=true, backbone=ViT-L/14+2023.03 | 61.6 | 37.1 | 71.5 | — | |
| OpenAI CLIP-L/14zero-shot=true, backbone=ViT-L/142023.03 | 61 | 36.5 | 71.1 | — | |
| Random Ranking2015.04 | 60 | 10 | 110 | 631 | |
| OpenAI CLIP-B/16zero-shot=true, backbone=ViT-B/162023.03 | 58.4 | 33.1 | 69 | — | |
| Random Ranking2015.04 | 50 | 10 | 100 | 500 |