Image-Text Matching on What’sUp
26.2AURCPAPSP-S
Evaluation Results
| Method | Links | |
|---|---|---|
| PAPSP-SModel Scale=Small, Memory Augmentation=false2026.01 | 26.2 | |
| PAPSPsize=S2026.01 | 26.2 | |
| PAPSPBackbone=SigLIP B/162026.01 | 26.2 | |
| ReCoVERR2026.01 | 24.9 | |
| Aesthetics2026.01 | 24.1 | |
| GenEval2026.01 | 24 | |
| VQAScore2026.01 | 23.6 | |
| SeeTRUE2026.01 | 23.6 | |
| VQAScore2026.01 | 23.6 | |
| SeeTRUE2026.01 | 23.6 | |
| LYP2026.01 | 22.8 | |
| MA-PAPSP-SModel Scale=Small, Memory Augmentation=true2026.01 | 21.4 | |
| MA-PAPSPsize=S2026.01 | 21.4 | |
| MA-PAPSPBackbone=SigLIP B/162026.01 | 21.4 | |
| Qwen-VL-7BMethod Setup=LVLM-PAPSP, Backbone Model=Qwen-VL-7B2026.01 | 20.5 | |
| PAPSP-LModel Scale=Large, Memory Augmentation=false2026.01 | 20.2 | |
| PAPSPsize=L2026.01 | 20.2 | |
| PAPSPBackbone=SigLIP SO-400M2026.01 | 20.2 | |
| Qwen-VL-72BMethod Setup=LVLM-PAPSP, Backbone Model=Qwen-VL-72B2026.01 | 19.9 | |
| MA-PAPSP-LModel Scale=Large, Memory Augmentation=true2026.01 | 19.2 | |
| MA-PAPSPMethod Setup=MA-PAPSP, Backbone Model=SigLIP SO-400M2026.01 | 19.2 | |
| MA-PAPSPsize=L2026.01 | 19.2 | |
| MA-PAPSPBackbone=SigLIP SO-400M2026.01 | 19.2 |