Twitter user under the nickname StepnEurope used new generative fill in Photoshop to add context to popular memes. To do this, he enlarged the images and used a fill with a neural network, filling the space around the original image.
The white frame on the photo marks the area that is visible in the original meme. Everything around her was added by the neural network.
1. A girl in a child's chair (she is also a girl with teeth).
2. "Think about it."
3. "Probably thinking about his women."
4. Women yelling at a cat (it turned out that it wasn't really a cat).
5. Harold hiding the pain.
6. "Women!".
7. "Change my mind."
8. The same car at the fork.
9. The girl is a disaster.
10. The guy who looked the wrong way.
In the comments, they note that it turned out too well: the frames look like real originals of photos, not generated ones. And what memes would you like to see in an expanded format?
