Conditional Image Generation on ImageNet 100
11.747FIDMoEMA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MoEMACondition representation=CLIP text encoder2026.07 | 11.747 | 1.338 | — | |
| StaticMA2026.07 | 11.802 | 1.317 | 0.5 | |
| BestSingle2026.07 | 13.021 | 1.964 | 9.8 | |
| SimpleAvg2026.07 | 17.889 | 4.871 | 34.3 |