OpenAI has added a new “outpainting” function to its DALL-E text-to-image AI model that allows the system to generate new visuals that enlarge the borders of a given image.
In the example above you can see how DALL-E, with the help of human prompting, “imagines” what is outside the frame of Johannes Vermeer’s portrait “The Girl with a Pearl Earring”. Note how, even from the limited information provided by the portrait, the system is able to match Vermeer’s style, mimicking the shadows and highlights of the original.
In the timelapse below, you can also see how the artist in charge, August Kamp, had to enlarge the image in small sections at a time, often redoing DALL-E generations in order to achieve the result it wanted. wished. What is not seen in this video but certainly worth highlighting is the fact that the system does not generate these extensions on its own. As with any text-to-image AI, the model requires humans to describe the new visuals.
The Outpainting feature can be used to extend the original content, but of course many DALL-E users have played around with this feature to see what lies outside the frame of famous images. (Scroll down for my absolute favorite example…)
From a broader perspective, outpainting doesn’t really extend the core functionality of text-image AI systems, but it does show how OpenAI will likely position itself in the growing market for such systems: by making user-friendliness a key argument for customers.
Many text-to-image AI models can perform the same essential function as paint, but, like DALL-E itself before this update, it required quite a bit of manual tinkering. Making painting as easy as possible will help DALL-E differentiate itself from the growing competition from smaller but comparable systems like Midjourney and Stable Diffusion.
DALL-E itself is now available through a beta program, which currently has access to over one million users. Each beta user gets 50 free image builds in the first month, then 15 more uses each month thereafter. They can then purchase an additional 115 generations of images for $15.
In the meantime, however, the paint can be used to answer some of life’s greatest mysteries, like “what if the guy from Quaker Oats was a busty bartender?” Wonder no more: