Dreambooth is a method of fine-tuning your own text-to-image diffusion models which can then be used with Stable Diffusion. When training your own model, you’re required to upload several input images – these will serve as the base for your model. For example, if you want to create a model for your pet, you’d upload 5-20 images of your pet from different angles. You can read the full paper for Dreambooth on this page. Fortunately, a site called “getimg.ai” provides an easy way to access and use Dreambooth which is much more accessible and streamlined.
