We like series. Thematic image series. The editorial team has gone crazy with a series that shows the multiple hairstyles for women. Terms like pixie, bob, lob, shag cut, wolf cut… what are we talking about? Well, just hairstyles.
When we create an image to illustrate an article or a story and the AI generates a woman, it is the AI itself that decides what she will look like unless we indicate all her characteristics. Age, skin color, eyes, mouth, nose… and hairstyle. For the hairstyle, we surely end up with few options: short hair, long hair, and at most the color of the same.
With patience and batch generation, we’ll see that there are dozens and dozens of hairdressing styles that especially men can’t even imagine:
Pixie, bob, mini bob, carré bob, genderless, mullet, bowl cut, garçon… and the list goes on and on.
Stable Diffusion is quite good at generating the style. A batch in which we specify a variable phrase like {pixie|bob|mullet….} within the prompt will generate the same woman with similar characteristics but with different hair.
MidJourney plays in another league and will generate completely different images with the hairstyle we assign. It also allows using limited variants based on the type of subscription. The maximum number in the worst case is four and it indicates something different: {pixie, bob, mullet…}, instead of using the separator “|”, it uses a comma.
In the magazine, we will see all these possibilities in more depth. For now, some samples made on the fly. Can you recognize these cuts?
