`` unprompt.ai '' that allows you to search for images and prompts that can be generated by Stable Diffusion and Midjourney of image generation AI from photos and illustrations
Image generation AI such as Stable Diffusion and Midjourney automatically generates an image according to the content by entering a character string called a prompt. ' unprompt.ai ' is a search engine that displays a similar image and its prompt by loading an image that is similar to the image you want to generate. It is useful when you want to generate an image.
Search 35+ Million AI Art Prompts
When you access unprompt.ai, it looks like this. Click the camera icon to the right of the entry field at the top of the screen.
Torottama Cheese Teriyaki Burger ~ Using Hokkaido Gouda Cheese ~ '.
Explorer will start, so select a photo or illustration that will be the original image. This time, I selected the following image of '
Then, the image of the hamburger was displayed as a search result like this. Click on the top left image.
(CFG scale ), Steps (number of steps generated), Sampling Method (sampler), and Seed (seed value) are displayed. If you enter this prompt and parameters as they are in the target model, the same image will be generated.
Then, in the 'Main Prompt' part, there is a prompt for generation, and in 'Parameters' below it, Model Name (model name of image generation AI), App (application), Image Size (image size), Model Version (model version), CFG Scale
It is like this if it is an image of another burger. Looking at the model name, it is 'Midjourney'. Since Midjourney is generated by the chat tool Discord , parameters such as CFG scale and number of generation steps are unknown.
Princess and Gamer '.
This time, I tried reading the illustration of the manga '
Then, the illustration was displayed as the search result as follows.
Niji Journey, which specialized Midjourney for two-dimensional moe illustrations.
The illustration is very complete, so when I checked the generative model, it was an illustration generated by