TIGER, a text-to-image GAN that stacks CLIP and DINO representations in the discriminator, reports FID 5.48 on COCO and 9.38 on CUB, surpassing prior GAN and some diffusion baselines.
Vector quantized diffusion model for text- to-image synthesis
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Text-to-Image GAN with Pretrained Representations
TIGER, a text-to-image GAN that stacks CLIP and DINO representations in the discriminator, reports FID 5.48 on COCO and 9.38 on CUB, surpassing prior GAN and some diffusion baselines.