When I add a specific artist to the prompt, copyrighted characters often don't generate properly. However, I'm getting tired of this model's default style, so I'm wondering if there's anything I can do about it.
What artists where you trying to use? I have a list of ones that work in NovelAI.
When I add a specific artist to the prompt, copyrighted characters often don't generate properly. However, I'm getting tired of this model's default style, so I'm wondering if there's anything I can do about it.
Your image is very clean. Did you train using screenshots of the anime? I have tried making loras from 640 x 480 sized images, since I cannot get any better quality from those DVDs, but the outputs tend to have noise or melted fingers. Mind telling me how you pulled this off?
I would not know, you would have to ask casualkaze.
From my online experience, many people who argue for 'pushing borders' would use these works as a proof that the alien, the perverse, even the ugly fascinates people more than the common and the beautiful. But I'm amazed at your ability to prompt the human out of the inhuman, the familiar out of the alien, the beauty out of the sublime. The material you trained your model on is your choice. Even as some of your (your AI's) works won't pass the test of the categorical imperative and thus remain on the other side of my red line, you (your AI) make many things agreeable for my taste that generally disagree with me. Ignoring the taboo you demonstrate the simple truth that youth per se is not a threat of ruin but a promise of growth. What makes these works light and agreeable is your disinterest in power. The cornerstone of both humanism and aesthetics is impartial admiration.
If you think something is beautiful, then its opposite becomes ugly. In other words, making something beautiful is also making something ugly.
This original commentary already comes with the English translation. In this case, there is no need to translate it, but is it enough to remove the commentary request tag?
This original commentary already comes with the English translation. In this case, there is no need to translate it, but is it enough to remove the commentary request tag?
@stifffits Referencing Kant makes me feel you might've been stoned while writing this, but honestly, I appreciate it :D
I'm mostly trying to make images that look good. Contrary to what some people seem to think, I'm not into shock content. I'm here trying to have fun while looking at nicely rendered images of anime girls, not necessarily to fap.
I do admit that I'm an attention whore though, so that's why I post unusual stuff sometimes. But my goal has never been to make people gag from looking at them, it's more about making people go "hey, that's kinda cool." So I'm glad you took the time to write this feelsgoodman.jpg
I don't agree with your categorical imperative assessment btw. Unless you meant that it would be bad if people were *doing* the things that are in my images. Then yeah, let's refrain from making that real.
From my online experience, many people who argue for 'pushing borders' would use these works as a proof that the alien, the perverse, even the ugly fascinates people more than the common and the beautiful. But I'm amazed at your ability to prompt the human out of the inhuman, the familiar out of the alien, the beauty out of the sublime. The material you trained your model on is your choice. Even as some of your (your AI's) works won't pass the test of the categorical imperative and thus remain on the other side of my red line, you (your AI) make many things agreeable for my taste that generally disagree with me. Ignoring the taboo you demonstrate the simple truth that youth per se is not a threat of ruin but a promise of growth. What makes these works light and agreeable is your disinterest in power. The cornerstone of both humanism and aesthetics is impartial admiration.
Video created using my previous work and model Auriel (that is what I call the artwork, I use her to test AI image and video creators for how far they will go. Now that I understand the actual model tag more...) in KlingAI. Need to go back and work on the hair. movement.
Your image is very clean. Did you train using screenshots of the anime? I have tried making loras from 640 x 480 sized images, since I cannot get any better quality from those DVDs, but the outputs tend to have noise or melted fingers. Mind telling me how you pulled this off?
Actually, there's a ton of obvious artifacts that I didn't notice upon original inspection. I wouldn't mind seeing this image unapproved and/or deleted so I can try again.
@bahahamut Thanks! I tried making another pixiv but barely anyone viewed my images so I stopped uploading. I still have antlers.cc if you're interested. I sometimes even update it.
Just my two cents, which doesn't mean anything. This isn't directed at the artist but the community as a whole. I'm wondering why we don't keep censored work on their respective websites that require it (ie. px), and have only uncensored everywhere else that isn't mandated by antiquated laws? It only supports censorship when people post their censored works on non censorship requiring websites, like this. Even if you're posting within your own country, if you can only post censored works you should stick with censored-only websites where people expect it, especially if the censorship is obvious and gaudy. Otherwise, it weakens the position of uncensored communities when censored work is posted, and it looks like a kid in a group of adults sharing their 'not yet quite adult' pictures...
