can we have normal sized version? also a bit more variety for the face type/angle make up and expression.(and lighting should be better cause it uses out-door background.) How many per batch you generate and pick the better ones from them? During my brief research of prompting and trying to posing the result, it’s really hit and miss.(even for the “hit” one you probably still need to hand tweak them to make them look nicer.)
I haven’t played around with the newer iterations, kinda dropped it and waiting for direct-ml to become more mature. It’s a useful iteration tool for me though to generate texture patterns then pipe to program that make them tileable.
Not really… sure you can run a1111/comfy locally but you’d still be bound by the terms of the Checkpoint finetunes and any LoRa’s you’ve used, and a lot of those are really sketchy about listing their sources.
Plenty of orgs get caught when they don’t even pay for the the software on their computers. I wouldn’t be at all surprised to find that the same entities not willing to pay a person to make artwork try to skip out on paying a business to generate it
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
No game suggestions, friend requests, surveys, or begging.
No Let’s Plays, streams, highlight reels/montages, random videos or shorts.
No off-topic posts/comments, within reason.
Use the original source, no clickbait titles, no duplicates.
(Submissions should be from the original source if possible, unless from paywalled or non-english sources.
If the title is clickbait or lacks context you may lightly edit the title.)
Up vote but the boob (to our right) looks kinda weird.
can we have normal sized version? also a bit more variety for the face type/angle make up and expression.(and lighting should be better cause it uses out-door background.) How many per batch you generate and pick the better ones from them? During my brief research of prompting and trying to posing the result, it’s really hit and miss.(even for the “hit” one you probably still need to hand tweak them to make them look nicer.)
I haven’t played around with the newer iterations, kinda dropped it and waiting for direct-ml to become more mature. It’s a useful iteration tool for me though to generate texture patterns then pipe to program that make them tileable.
If they are too cheap to pay original artists, why should one pay the original developer to play?
So they don’t have to replace even more with AI
we ddin’t start replacing
Ahoy!
Ah so the artwork isnt copyrightable and anyone can use it.
Whose AI system are they using, and what did they pay for it? A lot of them do have clauses prohibiting commercial use of generated images
It’s entirely possible they used Stable Diffusion locally, which means anything goes as long as it’s legal and is not harmful to any individuals.
Not really… sure you can run a1111/comfy locally but you’d still be bound by the terms of the Checkpoint finetunes and any LoRa’s you’ve used, and a lot of those are really sketchy about listing their sources.
It’s called signing a contract and paying something. It’s used a lot nowadays.
Plenty of orgs get caught when they don’t even pay for the the software on their computers. I wouldn’t be at all surprised to find that the same entities not willing to pay a person to make artwork try to skip out on paying a business to generate it