The creation of images with artificial intelligence and text has given rise to what is known as "prompt engineering", which allows us to guide the generation process and obtain better results. However, it is not the only option available: the Img2Img mode enables injecting a preexisting image as reference, and that brings us to Stable Doodle, another free feature of the ClipDrop platform that transforms doodles and freehand drawings into complete images.
In mid-April we talked about Animated Drawings, an open source project published by Meta that helps us animate basic drawings and other doodles using artificial intelligence. But we cannot lose sight of something essential, and that is that those same drawings can modify the behavior of an algorithm when generating images. The concept of Img2Img is very useful when the initial results are not satisfactory, and the ClipDrop platform places it within reach with Stable Doodle.
Turn simple drawings into images with artificial intelligence
All you have to do is draw on the canvas, add a basic prompt, choose a style (this is optional), and click the Generate button. As always happens in ClipDrop, Stable Doodle presents its results in a 2x2 grid. The first square shows the original drawing, and the other three can be replaced/updated with additional clicks on Generate. If a particular image has what you are looking for, remember to download its HD version, which has a final size of 2048 x 2048.
Needless to say, the margin of error is enormous in Stable Doodle, and you will have to modify your doodles several times to get the desired image. At the same time, drawing with the mouse is not a very comfortable option, and all images include a watermark in the free version (though that is easily resolved). To finish, Stable Doodle works best with a ClipDrop account, or signing in with alternative credentials (Google, Apple, Facebook). Give it a try!
Official website: Click here
