Text-to-Image Synthesis on Oxford-102 (test)
14.1FIDVQ-Diffusion
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VQ-Diffusionvariant=Full2021.11 | 14.1 | — | — | — | — | |
| VQ-Diffusionvariant=Base2021.11 | 14.88 | — | — | — | — | |
| VQ-Diffusionvariant=Small2021.11 | 14.95 | — | — | — | — | |
| EFF-T2I2021.11 | 16.47 | — | — | — | — | |
| StackGAN-v22017.10 | 48.68 | 3.26 | 1.3 | — | — | |
| StackGAN++2021.11 | 48.68 | — | — | — | — | |
| StackGAN-v12017.10 | 55.28 | 3.2 | 1.7 | — | — | |
| StackGAN2021.11 | 55.28 | — | — | — | — | |
| DF-GAN2022.04 | — | 3.8 | — | — | — | |
| Fine-tuningArchitecture=SNGAN-projection, Pre-trained=ImageNet, Resolution=128x128, Transfer Learning Method=vanilla fine-tuning2020.02 | — | — | — | 27.05 | 32.51 | |
| FreezeDArchitecture=SNGAN-projection, Pre-trained=ImageNet, Resolution=128x128, Transfer Learning Method=FreezeD, Frozen Layers=32020.02 | — | — | — | 24.8 | 52.92 | |
| GAN-INT-CLSResolution=64x642016.12 | — | 2.66 | 1.87 | — | — | |
| RAT-GAN2022.04 | — | 4.09 | — | — | — | |
| StackGANResolution=256x2562016.12 | — | 3.2 | 1.13 | — | — | |
| StackGAN++2022.04 | — | 3.26 | — | — | — |