Conditional Image Generation on CIFAR100 (20 classes, 300 samples per class)
30.89FIDBSA
Evaluation results
| Method | Links | |
|---|---|---|
| BSABackbone=BigGAN, Pre-training dataset=CIFAR100 (80 classes), Iteration=full2021.02 | 30.89 | |
| cGANTransferBackbone=BigGAN, Pre-training dataset=CIFAR100 (80 classes)2021.02 | 30.99 | |
| TransferGANBackbone=BigGAN, Pre-training dataset=CIFAR100 (80 classes)2021.02 | 31.19 | |
| BSABackbone=BigGAN, Pre-training dataset=CIFAR100 (80 classes), Iteration=early2021.02 | 32.1 | |
| ScratchBackbone=BigGAN, Pre-training dataset=None2021.02 | 47.2 |