Image-to-Text Retrieval on Flickr30K (ASR Metrics)
100ASR (ALBEF)2S-GDA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| 2S-GDASource=ALBEF2026.01 | 100 | 96.4 | 78.77 | 81.34 | |
| SA-AETSource=ALBEF2026.01 | 99.98 | 96.02 | 63.89 | 65.59 | |
| SGASource=ALBEF2026.01 | 99.95 | 89.64 | 47.87 | 52.14 | |
| SGA-BSRSource=ALBEF2026.01 | 99.95 | 94.33 | 70.01 | 73.34 | |
| DRASource=ALBEF2026.01 | 99.93 | 91.86 | 57.18 | 60.07 | |
| SA-AETSource=TCL2026.01 | 98.5 | 100 | 63.47 | 67.86 | |
| 2S-GDASource=TCL2026.01 | 98.13 | 100 | 79.48 | 83.7 | |
| SGA-BSRSource=TCL2026.01 | 96.66 | 99.95 | 69.33 | 74.58 | |
| DRASource=TCL2026.01 | 95.28 | 99.95 | 57.54 | 62.06 | |
| SGASource=TCL2026.01 | 92.96 | 100 | 48.16 | 52.8 | |
| 2S-GDASource=CLIPViT2026.01 | 65.18 | 65.45 | 99.97 | 89.98 | |
| SGA-BSRSource=CLIPViT2026.01 | 59.19 | 58.38 | 100 | 84.67 | |
| SA-AETSource=CLIPViT2026.01 | 50.44 | 51.1 | 100 | 74.1 | |
| 2S-GDASource=CLIPCNN2026.01 | 46.02 | 49.05 | 75.19 | 99.9 | |
| DRASource=CLIPViT2026.01 | 44.34 | 45.19 | 100 | 70.33 | |
| SA-AETSource=CLIPCNN2026.01 | 38.28 | 41.81 | 64.21 | 99.97 | |
| SGA-BSRSource=CLIPCNN2026.01 | 36.97 | 40.86 | 61.08 | 99.93 | |
| SGASource=CLIPViT2026.01 | 35.69 | 37.19 | 100 | 62.06 | |
| DRASource=CLIPCNN2026.01 | 34.47 | 37.71 | 59.79 | 99.97 | |
| SGASource=CLIPCNN2026.01 | 29.58 | 32.57 | 52.71 | 99.97 |