Image Generation on TEXTBINDEVAL
0.64CLIP Sim (T1)MIM
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| MIMMedium=Language Description2023.09 | 0.64 | 0.645 | 0.673 | 0.712 | 0.72 | 0.726 | 139.46 | |
| SDXL2023.09 | 0.612 | 0.599 | 0.608 | 0.712 | 0.735 | 0.735 | 144.76 | |
| GILL2023.09 | 0.569 | 0.55 | 0.53 | 0.712 | 0.734 | 0.742 | 158.64 | |
| Q-FormerMedium=Prompt Tokens2023.09 | 0.566 | 0.571 | 0.606 | 0.718 | 0.727 | 0.732 | 153.23 | |
| Q-FormerMedium=Standard2023.09 | 0.558 | 0.568 | 0.592 | 0.717 | 0.728 | 0.729 | 155.01 |