Image Generation On Imagenet 128X128
評価指標
FID
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | FID |
|---|---|
| stylegan-xl-scaling-stylegan-to-large-diverse | 1.81 |
| large-scale-gan-training-for-high-fidelity | 8.7 |
| normalizing-flows-are-capable-generative | 5.03 |
| improved-transformer-for-high-resolution-gans | 30.83 |
| diffusion-models-beat-gans-on-image-synthesis | 2.97 |
| partition-guided-gans | 21.73 |
| regularizing-generative-adversarial-networks | 6.54 |
| cascaded-diffusion-models-for-high-fidelity | 3.52 |
| large-scale-gan-training-for-high-fidelity | 5.7 |
| polynomial-implicit-neural-representations | 2.08 |
| bigroc-boosting-image-generation-via-a-robust | 2.77 |
| モデル 12 | 1.75 |
| simpler-diffusion-sid2-1-5-fid-on-imagenet512 | 1.26 |
| self-supervised-generative-adversarial | 43.87 |
| consistency-regularization-for-generative-1 | 6.66 |
| モデル 16 | 9.92 |
| モデル 17 | 20.76 |
| optimal-budgeted-rejection-sampling-for | 11.65 |
| scalable-adaptive-computation-for-iterative | 2.75 |
| disco-diff-enhancing-continuous-diffusion | 1.73 |
| pagoda-progressive-growing-of-a-one-step | 1.48 |
| bigroc-boosting-image-generation-via-a-robust | 2.53 |
| differentiable-augmentation-for-data | 6.8 |