3 research papers to understand text-to-image synthesis models better.
1. The original latent diffusion paper @LMU_Muenchen
2. @GoogleAI’s Parti
3. The original VQGAN+CLIP paper
A thread 🧵👇
1. The original latent diffusion paper @LMU_Muenchen
2. @GoogleAI’s Parti
3. The original VQGAN+CLIP paper
A thread 🧵👇
1. The original latent diffusion paper
Latent diffusion is one of the most popular techniques for text-to-image synthesis models.
Paper: arxiv.org
Code: github.com
Latent diffusion is one of the most popular techniques for text-to-image synthesis models.
Paper: arxiv.org
Code: github.com
2. Google AI’s Parti
Google’s attempt to provide an alternative to text-to-image diffusion methods by using autoregressive models.
Paper: arxiv.org
Code: github.com
Google’s attempt to provide an alternative to text-to-image diffusion methods by using autoregressive models.
Paper: arxiv.org
Code: github.com
3. The original VQGAN+CLIP paper
Combines GAN and CLIP to generate high-quality images from textual descriptions.
Paper: arxiv.org
Code: github.com
Combines GAN and CLIP to generate high-quality images from textual descriptions.
Paper: arxiv.org
Code: github.com
Loading suggestions...