I am struggling to generate Disney Pixar-style images using DALL-E 3. I am including this in all my prompts: “The character in the image should be rendered in a 3D animated style, similar to modern Disney or Pixar films. The animation has a high degree of realism in terms of texture and lighting but maintains a cartoony, exaggerated cuteness.” However, the results don’t match the typical Disney Pixar style. Any leads?
I want to generate images like this:
But DALLE always ends up generating images like this:
{
“prompt”: “A three-year-old child playing with a stacking ring toy in his bedroom. The child is smiling and focused on the toy, which is colorful and made of different-sized rings. The bedroom is cluttered with toys in the background, including stuffed animals, cars, and building blocks. The style is 3D computer-rendered children’s movie animation with vibrant colors and detailed textures.”,
}
My advice would be to find a different approach than to mention Disney, Pixar, etc. (Much like @_j has done). It appears to be a behaviour caused by copyright issues, hence the need to avoid specifying the copyright holder directly (in my experience anyway).
edit: An image like this one - not generated by AI - you should not ask the AI to reproduce striking facsimiles that could only originate from the mass ingestion of copyrighted works without license…
Thanks for your help. I am getting much better results now. Additionally, could you help me try a prompt related to hair? Newborns and young babies typically don’t have that much hair or such neatly organised hair. But I am getting images with very fine and organized hairs. Thanks!
