Image Generation On Ffhq 256 X 256
评估指标
FID
评测结果
各个模型在此基准测试上的表现结果
比较表格
| 模型名称 | FID |
|---|---|
| generative-adversarial-transformers | 7.42 |
| a-u-net-based-discriminator-for-generative | 11.48 |
| clr-gan-improving-gans-stability-and-quality | 3.37 |
| stylenat-giving-each-head-a-new-perspective | 2.05 |
| exposing-flaws-of-generative-model-evaluation-1 | 5.30 |
| stylegan-xl-scaling-stylegan-to-large-diverse | - |
| 模型 7 | 8.11 |
| adversarial-generation-of-continuous-images | 4.95 |
| anycost-gans-for-interactive-image-synthesis | 3.35 |
| the-gan-is-dead-long-live-the-gan-a-modern | 2.75 |
| regularizing-generative-adversarial-networks | 3.49 |
| efficient-vdvae-less-is-more | 34.88 |
| exposing-flaws-of-generative-model-evaluation-1 | 8.11 |
| unleashing-transformers-parallel-token | - |
| data-efficient-instance-generation-from | - |
| improved-transformer-for-high-resolution-gans | 3.06 |
| improved-transformer-for-high-resolution-gans | 2.58 |
| very-deep-vaes-generalize-autoregressive-1 | - |
| styleswin-transformer-based-gan-for-high-1 | 2.81 |
| exposing-flaws-of-generative-model-evaluation-1 | 3.46 |
| score-matching-model-for-unbounded-data-score-1 | 5.54 |
| data-efficient-instance-generation-from | 3.31 |
| polynomial-implicit-neural-representations | 2.72 |
| stylenat-giving-each-head-a-new-perspective | - |
| stylegan-xl-scaling-stylegan-to-large-diverse | 2.19 |
| gotta-go-fast-when-generating-data-with-score | 15.67 |
| efficient-vdvae-less-is-more | - |
| exposing-flaws-of-generative-model-evaluation-1 | 9.02 |
| training-generative-adversarial-networks-with-2 | - |
| adversarially-slicing-generative-networks | 1.68 |
| training-generative-adversarial-networks-with-2 | 3.62 |
| exposing-flaws-of-generative-model-evaluation-1 | 34.88 |
| exposing-flaws-of-generative-model-evaluation-1 | 2.11 |
| projected-gans-converge-faster | 3.39 |
| a-u-net-based-discriminator-for-generative | 7.48 |
| gotta-go-fast-when-generating-data-with-score | 15.67 |
| compensation-sampling-for-improved | 2.57 |
| swagan-a-style-based-wavelet-driven | 5.22 |
| 模型 39 | - |
| nvae-a-deep-hierarchical-variational | - |
| exposing-flaws-of-generative-model-evaluation-1 | 2.89 |
| styleswin-transformer-based-gan-for-high-1 | - |
| exposing-flaws-of-generative-model-evaluation-1 | 2.26 |
| 模型 44 | - |
| exposing-flaws-of-generative-model-evaluation-1 | 4.29 |
| taming-transformers-for-high-resolution-image | 9.6 |
| unleashing-transformers-parallel-token | 6.11 |
| improved-transformer-for-high-resolution-gans | 2.95 |
| compositional-transformers-for-scene-1 | 7.77 |
| flow-matching-in-latent-space | 4.55 |
| image-generators-with-conditionally | 4.38 |