Have you noticed? The “realistic” or “realism” AI results for man/woman have apparent big heads (mostly even huge) compared to their own bodies. It’s like the camera is close to the face. Even the results in Google Search have such phenomena. The obvious ones are for Asian women. Why as such for perhaps even the different AI’s, why the same pattern from them?
Try searching for such now, try something like: “realistic AI Asian woman”, and you’ll even see ones with actual HUGE heads, most especially the “realistic” ones. Different AI programs and yet the same huge head pattern from them?
Any actual AI users, could you please share your experience too? Thanks in advance for any enlightenment.
I’ve also seen a lot of drawn art where artists like to make heads quite big that also label their art as ‘realistic’.
If you’re trying to generate something real looking, I’d go for a prompt like ‘selfie’ or ‘photograph’. Because usually photos of real people aren’t labelled ‘real’ or ‘realistic’
Gotta remember that the training dataset is just a bunch of images labelled (mostly) by humans
My guess is focal length. Most facial photos with tons of detail will be close up selfies with a short focal length, thus increasing size of the face proportionally.
Generally I think this has something to do with average age of model trained vs age of depicted person. Humans have large heads proportional to their bodies in youth and then averages out. On top of this another commenter mentioned art. So many anime/cartoon characters have larger heads to show the expressions and in turn show a youthful nature in these characters.
Another addition to this is, as another poster stated, the angles of these photos makes a larger head more apparent, probably because we associate larger heads with intelligence, and so poses have just developed to make that appearance.
Similar to “the average number of arms is less than 2 because there’s greater chances there will be people with zero or one arm, rather than an extra”. Your model will draw on its wealth of samples where the head is disproportionately larger, instead of proportionate - because it was probably fed more photos of hat age range and doesn’t know any different. You asking for AI to give you a person like Trump (78) will show that disparity in head size more than if you ask for Greta Thunberg (21)
I was the base model for all AI image generation and my ego is fucking huge.
I imagine that since there are so many examples of characters with large heads throughout art, especially cartoons, it considers oversized head an acceptable variable to play with. The more realistic the art form, the less variation you see, but there are still big headed realistic drawings and paintings.
Also so many head photos taken from an angle right in front of the face or a bit higher, distorting the shoulders in ways you don’t notice but which matters when the algorithm tries to splice everything together into an averaged model
deleted by creator
