Photographers and graphic designers have created an incredible set of AI images. The Simpsons Uses the latest text to image conversion technology.
Millie, known online as Princess Prompto, Petapixel She uses the latest artificial intelligence technologies (such as Midjourney v6 and Upscaler) to The Simpsons letter
“I wanted to try out the new features in the latest version of Midjourney,” Millie explains. The Simpsons’ I’m a fan, especially because they all have such a distinct look with very bright colors and unique accessories.”
methodology
Millie found screenshots for each character she wanted to recreate and used those images as prompts using the “image weight” parameter. She then described the screenshots to Midjourney as “accurately and succinctly” as she could.
The task of this exercise is The Simpsons“I didn’t want the characters in Heroines to be portrayed as yellow-skinned, I wanted people to see them as people they encounter in their everyday lives,” Millie says.
After defining the overall aesthetic, Millie used the same basic prompt for each character, refining specific details.
“Once all the characters were generated, we sent them to Magnific.ai artificial intelligence to add details like skin texture, hair, wrinkles, etc.,” she says.
“I generated all the characters in a week, mostly at night. I didn’t put any pressure on myself, as it was more of a fun exercise than anything else. Midjourney is addictive and you can spend a lot of time on it just for the pleasure of looking at the generated images.”
“Some characters were very easy to create, while others took much longer. For example, with the character of Smithers, it was difficult to accurately define his hairstyle and find the words to make it look similar,” she adds.
Millie noted that using “Marge Simpson” as the prompt for “Midjourney” would have resulted in a “cartoonish” result, so she humanized the Simpson family matriarch by describing her as “a mature woman with blue, curly beehive-style hair.”
“In my opinion, the most time-consuming part is the final image selection. You have to organize things and it can get complicated because the artificial intelligence provides you with hundreds of images in just a few minutes,” she added.
“In these cases, you have to show the discipline to stick to the aesthetic and idea you were aiming for at the start.”
Surprising results
Millie’s eye-catching AI image caused a huge stir on Facebook, garnering nearly 100,000 responses.
“The reactions reveal the curiosity, distrust, contempt, and sometimes fear that generative AI provokes,” Milley notes. “The numbers also confirm the deep affection people have for it. The Simpsons.”
Millie is also very impressed with the latest Mid Journey. Petapixel The results were “much more accurate and faithful to the initial prompts.”
Millie has been practising “fine art photography” for around 20 years, and says she enjoys taking costumed portraits in a studio, then retouching them on a computer and letting her creativity run wild.
“New tools and new forms of creativity always fascinate and intrigue me in any form, so when generative artificial intelligence first emerged among the general public last summer, I was immediately fascinated and amazed by the possibilities it opened up for me,” she says.
“I started out with Disco Diffusion, then Stable Diffusion, and I also tested Dall-E, but my preference and favorite remains Midjourney in terms of the way it produces such rich, detailed images.”
“My ultimate goal is to fuse what I’ve personally created with artificial intelligence, or to train my own artificial intelligence using a local version of Stable Diffusion, for example,” she added.
To see more of Milie’s photography work, check out her website and Instagram , and to follow her experiments with AI, check out her Facebook page .
Image credits: Princess Prompt
