Fetching the paper…

Zero-shot spatial layout conditioning for text-to-image diffusion models · Around