In the next sections we’ll describe how text is incorporated in the process in order to control what type of image the model generates. So if we deploy this model, it would generate great looking images, but we’d have no way of controlling if it’s an image of a pyramid or a cat or anything else. Note that the diffusion process we described so far generates images without using any text data. Now that you have this intuition of diffusion, you know the main components of not only Stable Diffusion, but also Dall-E 2 and Google’s Imagen. This concludes the description of image generation by diffusion models mostly as described in Denoising Diffusion Probabilistic Models.
0 kommentar(er)
