Text Retrieval on Flickr30K (test) using ASR Metrics
100R@1 (ASR)SGA(LI)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SGA(LI)Source=ALBEF, Target Model=ALBEF2026.03 | 100 | 100 | 100 | |
| SGA(LI)+SIASource=ALBEF, Target Model=ALBEF2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)Source=ALBEF, Target Model=ALBEF2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)+SIASource=ALBEF, Target Model=ALBEF2026.03 | 100 | 100 | 100 | |
| SADCASource=ALBEF, Target Model=ALBEF2026.03 | 100 | 100 | 100 | |
| SGASource=TCL, Target Model=TCL2026.03 | 100 | 100 | 99.9 | |
| SGA+SIASource=TCL, Target Model=TCL2026.03 | 100 | 99.7 | 99.6 | |
| SGA(LI)Source=TCL, Target Model=TCL2026.03 | 100 | 100 | 100 | |
| SGA(LI)+SIASource=TCL, Target Model=TCL2026.03 | 100 | 99.9 | 99.9 | |
| DRASource=TCL, Target Model=TCL2026.03 | 100 | 100 | 99.8 | |
| SA-AETSource=TCL, Target Model=TCL2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)Source=TCL, Target Model=TCL2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)+SIASource=TCL, Target Model=TCL2026.03 | 100 | 100 | 100 | |
| SADCASource=TCL, Target Model=TCL2026.03 | 100 | 100 | 100 | |
| SGASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 99.9 | |
| SGA(LI)Source=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SGA(LI)+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| DRASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SA-AETSource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)Source=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SADCASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | 100 | 100 | |
| SGA(LI)+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | 99.68 | 99.38 | |
| SA-AETSource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | 100 | 99.9 | |
| SA-AET(LI)Source=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | 100 | 100 | |
| SA-AET(LI)+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | 99.79 | 99.79 | |
| SADCASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | 100 | 100 | |
| SGA(LI)+SIASource=ALBEF, Target Model=ALBEF2026.03 | 100 | — | — | |
| SA-AET(LI)+SIASource=ALBEF, Target Model=ALBEF2026.03 | 100 | — | — | |
| SADCASource=ALBEF, Target Model=ALBEF2026.03 | 100 | — | — | |
| SGASource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| SGA(LI)+SIASource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| DRASource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| SA-AETSource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| SA-AET(LI)+SIASource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| SADCASource=TCL, Target Model=TCL2026.03 | 100 | — | — | |
| SGASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| SGA(LI)+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| DRASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| SA-AETSource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| SA-AET(LI)+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| SADCASource=CLIPViT, Target Model=CLIPViT2026.03 | 100 | — | — | |
| SGA(LI)+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | — | — | |
| SA-AETSource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | — | — | |
| SA-AET(LI)+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | — | — | |
| SADCASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 100 | — | — | |
| SA-AET(LI)+SIASource=TCL, Target Model=ALBEF2026.03 | 99.95 | 99.9 | 99.8 | |
| SA-AET(LI)+SIASource=TCL, Target Model=ALBEF2026.03 | 99.95 | — | — | |
| SGASource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | 99.7 | 99.7 | |
| DRASource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | 99.7 | 99.7 | |
| SA-AETSource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | 99.8 | 99.8 | |
| SGA(LI)+SIASource=TCL, Target Model=ALBEF2026.03 | 99.9 | 99.7 | 99.6 | |
| SGASource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | — | — | |
| DRASource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | — | — | |
| SA-AETSource=ALBEF, Target Model=ALBEF2026.03 | 99.9 | — | — | |
| SGA(LI)+SIASource=TCL, Target Model=ALBEF2026.03 | 99.9 | — | — | |
| SGA+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 99.88 | 99.48 | 99.09 | |
| SGASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.87 | 99.47 | 99.07 | |
| SGA(LI)Source=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.87 | 99.89 | 99.9 | |
| DRASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.87 | 25.13 | 99.38 | |
| SGASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.87 | — | — | |
| DRASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.87 | — | — | |
| SGA+SIASource=ALBEF, Target Model=ALBEF2026.03 | 99.79 | 99.5 | 99.2 | |
| SA-AET+SIASource=TCL, Target Model=TCL2026.03 | 99.79 | 99.6 | 99.2 | |
| SA-AET+SIASource=CLIPViT, Target Model=CLIPViT2026.03 | 99.63 | 99.48 | 98.98 | |
| SGA+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.62 | 97.99 | 96.81 | |
| SA-AET+SIASource=ALBEF, Target Model=ALBEF2026.03 | 99.58 | 99.3 | 98.8 | |
| SA-AET(LI)+SIASource=ALBEF, Target Model=TCL2026.03 | 99.58 | 98.89 | 98.5 | |
| SADCASource=TCL, Target Model=ALBEF2026.03 | 99.58 | 99.2 | 99 | |
| SA-AET(LI)+SIASource=ALBEF, Target Model=TCL2026.03 | 99.58 | — | — | |
| SADCASource=TCL, Target Model=ALBEF2026.03 | 99.58 | — | — | |
| SA-AET+SIASource=CLIPCNN, Target Model=CLIPCNN2026.03 | 99.49 | 98.31 | 96.7 | |
| SGA(LI)+SIASource=ALBEF, Target Model=TCL2026.03 | 99.37 | 98.29 | 97.8 | |
| SA-AET(LI)Source=TCL, Target Model=ALBEF2026.03 | 99.37 | 98.6 | 98.2 | |
| SGA(LI)+SIASource=ALBEF, Target Model=TCL2026.03 | 99.37 | — | — | |
| PGDSource=TCL, Target Model=TCL2026.03 | 99.37 | — | — | |
| SA-AETSource=TCL, Target Model=ALBEF2026.03 | 98.85 | 96.79 | 95.5 | |
| SA-AETSource=TCL, Target Model=ALBEF2026.03 | 98.85 | — | — | |
| SA-AET(LI)Source=ALBEF, Target Model=TCL2026.03 | 98.63 | 97.29 | 96.09 | |
| SADCASource=ALBEF, Target Model=TCL2026.03 | 98.52 | 96.28 | 94.79 | |
| SADCASource=ALBEF, Target Model=TCL2026.03 | 98.52 | — | — | |
| SADCASource=CLIPViT, Target Model=CLIPCNN2026.03 | 97.9 | 92.81 | 89.91 | |
| SADCASource=CLIPViT, Target Model=CLIPCNN2026.03 | 97.9 | — | — | |
| SGA+SIASource=TCL, Target Model=ALBEF2026.03 | 97.18 | 94.19 | 91.9 | |
| SA-AET+SIASource=TCL, Target Model=ALBEF2026.03 | 97.08 | 93.79 | 92.6 | |
| Co-AttackSource=ALBEF, Target Model=ALBEF2026.03 | 97.08 | — | — | |
| SA-AETSource=ALBEF, Target Model=TCL2026.03 | 96.42 | 92.36 | 89.98 | |
| SA-AETSource=ALBEF, Target Model=TCL2026.03 | 96.42 | — | — | |
| Co-AttackSource=CLIPCNN, Target Model=CLIPCNN2026.03 | 95.91 | — | — | |
| SA-AET+SIASource=ALBEF, Target Model=TCL2026.03 | 95.21 | 90.01 | 85.97 | |
| DRASource=TCL, Target Model=ALBEF2026.03 | 95.2 | 91.28 | 88 | |
| DRASource=TCL, Target Model=ALBEF2026.03 | 95.2 | — | — | |
| SA-AET(LI)+SIASource=CLIPViT, Target Model=CLIPCNN2026.03 | 94.76 | 89.32 | 85.79 | |
| SA-AET(LI)+SIASource=CLIPViT, Target Model=CLIPCNN2026.03 | 94.76 | — | — | |
| SGA+SIASource=ALBEF, Target Model=TCL2026.03 | 94.1 | 88.74 | 85.07 | |
| PGDSource=ALBEF, Target Model=ALBEF2026.03 | 93.74 | — | — | |
| SGASource=TCL, Target Model=ALBEF2026.03 | 92.4 | 87.07 | 85.4 | |
| SGASource=TCL, Target Model=ALBEF2026.03 | 92.4 | — | — | |
| Co-AttackSource=TCL, Target Model=TCL2026.03 | 91.68 | — | — | |
| DRASource=ALBEF, Target Model=TCL2026.03 | 91.57 | 81.31 | 75.95 |