Text To Image Generation On Oxford 102
評価指標
FID
Inception score
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
モデル名 | FID | Inception score | Paper Title | Repository |
|---|---|---|---|---|
| StackGAN-v1 | 55.28 | 3.2 | StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks | |
| StackGAN-v2 | 48.68 | 3.26 | StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks | |
| VQ-Diffusion-S | 14.95 | - | Vector Quantized Diffusion Model for Text-to-Image Synthesis | |
| VQ-Diffusion-B | 14.88 | - | Vector Quantized Diffusion Model for Text-to-Image Synthesis | |
| RAT-Diffusion | 9.52 | 4.35 | Data Extrapolation for Text-to-image Generation on Small Datasets | - |
| VQ-Diffusion-F | 14.1 | - | Vector Quantized Diffusion Model for Text-to-Image Synthesis | |
| RAT-GAN | 16.04 | 4.09 | Recurrent Affine Transformation for Text-to-image Synthesis | |
| StackGAN | - | 3.2 | StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks |
0 of 8 row(s) selected.