Text-vision Alignment on Q-Eval-Video 100K (test)
0.607Instance SRCCQ-Eval-Score
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Q-Eval-Scorebackbone=Qwen2-VL-7B-Instruct, zero-shot=false2025.03 | 0.607 | 0.634 | 0.648 | 0.605 | |
| BLIP2Scorezero-shot=false2025.03 | 0.483 | 0.488 | 0.512 | 0.481 | |
| ImageRewardzero-shot=false2025.03 | 0.472 | 0.485 | 0.375 | 0.362 | |
| VQAScorezero-shot=true, official default weight=true2025.03 | 0.433 | 0.432 | 0.433 | 0.351 | |
| CLIPScorezero-shot=false2025.03 | 0.431 | 0.443 | 0.519 | 0.509 | |
| BLIP2Scorezero-shot=true, official default weight=true2025.03 | 0.218 | 0.25 | 0.295 | 0.296 | |
| CLIPScorezero-shot=true, official default weight=true2025.03 | 0.186 | 0.219 | 0.518 | 0.5 |