Comment on How to generate realistic people in Stable DiffusionparentComments−danielbln2yIf you prompt the model in the way that pushes it towards low quality amateur photography, the results tend to be more realistic and not glossy:"90s, single use camera, documentary, of anoffice worker in an open plan office, realistic, amateur photo, blurry"Results: https://imgur.com/a/GJLqYft−ethbr12yResults: https://imgur.com/a/GJLqYftCorporate accounts payable, Nina speaking. Just a moooment. https://m.youtube.com/watch?v=4s5yHUpumkY−hackable_sand2yIs he squinting or am I?−bastien22yonly a prompter would think that looks realistic−danielbln2yOh a "prompter" is it? No, I'm not a prompter, but if you want to interface with a GenAI model, a prompt is kind of the way to do it. No need to be a salty about it.
Comments
If you prompt the model in the way that pushes it towards low quality amateur photography, the results tend to be more realistic and not glossy:
"90s, single use camera, documentary, of anoffice worker in an open plan office, realistic, amateur photo, blurry"
Results: https://imgur.com/a/GJLqYft
Corporate accounts payable, Nina speaking. Just a moooment. https://m.youtube.com/watch?v=4s5yHUpumkY
Is he squinting or am I?
only a prompter would think that looks realistic
Oh a "prompter" is it? No, I'm not a prompter, but if you want to interface with a GenAI model, a prompt is kind of the way to do it. No need to be a salty about it.