Why don't indie devs use AI-generated images as art?

[Lagole Gon]

Somebody put a girl I drew into this image2image

Post booba

Spoiler

 

[Dexter]

I don't know how to use this zoomer shit.

But I want to challange this digital dragon. Somebody put a girl I drew into this image2image Moloch Machine:

I'm not particularly experienced with img2img, but basically how it works is if you have a Repo installed locally like this one or on one of those Colab services, you take your image and drag it into img2img. Then you have to describe the picture as best as you see fit. There's also an "interrogate" function to help you, I think it came up with something like "a black and white drawing of a woman's face with freckles on her face and a black collar" in this case.
From there you have to change the prompt to what you want it to look like and input your own style choices or artists, say "a color drawing of a red haired woman's face with freckles on her face and a small scar on her nose and a emerald necklace, by Charlie Bowater".

If you set your Denoise high you can immediately create however many variations of the image you like based on your composition, but that don't necessarily look like your Original image anymore, this is why it's especially potent on really shitty sketches since it can just wholly reinterpret them however it likes and potentially create something good from the get-go e.g.:

Spoiler

The settings you have to play with for img2img is CFG Scale (how close the image is supposed to stay to your prompt), try maybe somewhere around 12-15 depending or lower if you want it to be more creative and Denoising strength. The lower you set Denoise, the closer it'll resemble your Original image. Unfortunately if you input a B&W drawn sketch and want to change the image say to the point that it's colorized or similar, but that it'll largely keep your features, it won't do that if you don't roughly do it in Photoshop or whatever first, since it's too big a change at low Denoise.

Once you have your rough init image you can again describe it or what it's supposed to become, in this case something like "a drawing of a woman with red hair and green eyes, with a black shirt, wearing a necklace, scar across nose by Charlie Bowater". You can play around with CFG and Denoise till you Roll something you like, then you can take that image as new Init image and iterate, make any changes you want to it or the prompt and Roll again for infinite variations till you get something you like or start again/go a few steps back and try with a new art style/artists or by modifying the prompt a little:

Spoiler

Since I let it screw up the eyes in one of the earlier images by letting it fill too much, it propagated this in some of the follow-ups.
There are Face Fixing Algorithms included in the UI like GPFGAN or CodeFormer that you can enable, but they tend to remove small details and change expressions a bit:

 

[Dexter]

https://blogs.nvidia.com/blog/2022/09/23/3d-generative-ai-research-virtual-worlds/
https://github.com/nv-tlabs/GET3D

 

[Vermillion]

https://blogs.nvidia.com/blog/2022/09/23/3d-generative-ai-research-virtual-worlds/
https://github.com/nv-tlabs/GET3D

In the future there will only be "game" a subscription based service where AI generates "game" in "world" for your enjoyment retention. "Game" was specifically programmed under the guidance of multiple psychologists and diversity consultants to ensure that "game" remains within acceptable ethical parameters. In the year 20** it is illegal to play 3rd party unlicensed "game". It is illegal to jailbreak "game". Spending more than 8 hours per day on "game" will see a reduction in your social credit score. Playing less than 2.4 hours a day on "game" will see a reduction in your social credit score. Getting banned from "game" will see a reduction in your social credit score.
You rike?

 

[Darboven]

Can it generate animations for those 3d models?

 

[Dexter]

img2img of Pixel characters: https://www.reddit.com/r/StableDiff...e_fallout_2_characters_reimagined_by/ips52cz/

Someone using Inpainting/Outpainting:

 

[J1M]

Imagine paying $100k for an arts degree and watching someone who tries at this for a few weeks now knows more about various art styles and composition than you do. All because being an artist isn't gated by being able to hold a pencil anymore.

 

[Dexter]

Btw. last few days people have been using "Dreambooth" to insert new concepts directly into the model, while "Textual Inversion" can train on a bunch of images to produce a small external "embedding" file that can inject a special keyword with its own vectors at the text encoding stage, this actually extends the base model with a face of an actual person or a created character, object or activity that isn't in it and creates a new model file.

While it required ~38GB of VRAM to train just a few days ago, different people have been optimizing the code and now it's possible with 12.5GB:
https://github.com/ShivamShrirao/diffusers/tree/main/examples/dreambooth
https://github.com/huggingface/diffusers/pull/554#issuecomment-1259522002
https://github.com/XavierXiao/Dreambooth-Stable-Diffusion/issues/4

Edit: Now apparently less than 12GB VRAM: https://github.com/TheLastBen/fast-stable-diffusion

Presentation is YouTube cringe, but not a bad video showing what this can accomplish:

 

[J1M]

Wow, daily optimizations that cut memory usage in half! Someone must be replacing Python code with C.

 

[The_Sloth_Sleeps]

 

[The_Sloth_Sleeps]

Pretty much this thread.

 

[Dexter]

Various other AI-fied gaming stuff via img2img:

LOTS OF PICTURES

Can it generate animations for those 3d models?

It's not an Animation tool, so no. Also looks like the "Craiyon" of 3D model generation, there will probably be much more powerful tools in time. For Animation there are already proprietary solutions like: https://www.mixamo.com/#/ that can take a model, automatically rig it and apply basic animations like walking/running/jumping or whatever.

 

[Dexter]

FacecockAI experimenting with generating videos from text or still images: https://makeavideo.studio/

 

[Strig]

Imagine paying $100k for an arts degree and watching someone who tries at this for a few weeks now knows more about various art styles and composition than you do. All because being an artist isn't gated by being able to hold a pencil anymore.

I'm a fan of gatekeeping, especially skill-based gatekeeping, there's nothing good in lowering the standards. And you're wrong, most of the best results shown from any image generator were chosen by people with actual artistic background or at least some skill. Best proof for how wrong you are would be all these "comic books" made with AI generated images. It's clear that the authors think they made something special but it's also clear they don't even grasp the basics of sequential storytelling. At best they consider comic books to be a bunch of pretty pictures with repeating elements. Don't get me wrong, teaching yourself art without suffering (and paying for) the cancer of modern art education is great, but you still have to spend a considerable amount of time on it and there's no shortcut through AI generators.

 

[rusty_shackleford]

Pretty much this thread.

that's better than 99.9% of the shit arts degree holders produce though?

 

[rusty_shackleford]

couldn't imagine why these tools were created, anyways completely unrelated

 

[markec]

 

[rusty_shackleford]

artists who laughed at people who lost their jobs to foreign laborers:

 

[KateMicucci]

You're wasting your time ranting about art degrees. We already know that, with few exceptions, art schools are for lazy rich kids, not people who want to learn to draw and paint well. AI art doesn't hurt them because the selling prices of their "art" has more to do with their connections than their skills. You aren't sticking it to them and your schadenfreude is misplaced.

People who actually want to draw and paint are self-taught or take online courses, or they take private lessons from master artists in ateliers. They study the classics and learn the fundies. AI art does hurt them, by chopping off the lower rungs of the freelance jobs ladder. Developing artists who hoped to live off of their skill by selling commissions will have to compete with free AI programs. And third-world photobash monkies have another arrow in their quiver.

 

[rusty_shackleford]

You're wasting your time ranting about art degrees. We already know that, with few exceptions, art schools are for lazy rich kids, not people who want to learn to draw and paint well. AI art doesn't hurt them because the selling prices of their "art" has more to do with their connections than their skills. You aren't sticking it to them and your schadenfreude is misplaced.

People who actually want to draw and paint are self-taught or take online courses, or they take private lessons from master artists in ateliers. They study the classics and learn the fundies. AI art does hurt them, by chopping off the lower rungs of the freelance jobs ladder. Developing artists who hoped to live off of their skill by selling commissions will have to compete with free AI programs. And third-world photobash monkies have another arrow in their quiver.

can you explain how this is worse than chink sweatshops that crank out mass produced art constantly?

 

[rusty_shackleford]

I guess the reason it's bad is because instead of it being purely a tool used by globohomo to force a political agenda, now anyone can make anything they want?

 

[KateMicucci]

You're wasting your time ranting about art degrees. We already know that, with few exceptions, art schools are for lazy rich kids, not people who want to learn to draw and paint well. AI art doesn't hurt them because the selling prices of their "art" has more to do with their connections than their skills. You aren't sticking it to them and your schadenfreude is misplaced.

People who actually want to draw and paint are self-taught or take online courses, or they take private lessons from master artists in ateliers. They study the classics and learn the fundies. AI art does hurt them, by chopping off the lower rungs of the freelance jobs ladder. Developing artists who hoped to live off of their skill by selling commissions will have to compete with free AI programs. And third-world photobash monkies have another arrow in their quiver.

can you explain how this is worse than chink sweatshops that crank out mass produced art constantly?

It's cheaper, faster and there's no need to talk to chinks. I've never denied that third-world artists are a problem. It's the same principle as any other industry.

Anyone could make anything they wanted before art AI. They just had to learn art, which was a valuable, marketable skill. If you think that cheapening labor is bad for globohomo you're very confused.

 

[rusty_shackleford]

You're wasting your time ranting about art degrees. We already know that, with few exceptions, art schools are for lazy rich kids, not people who want to learn to draw and paint well. AI art doesn't hurt them because the selling prices of their "art" has more to do with their connections than their skills. You aren't sticking it to them and your schadenfreude is misplaced.

People who actually want to draw and paint are self-taught or take online courses, or they take private lessons from master artists in ateliers. They study the classics and learn the fundies. AI art does hurt them, by chopping off the lower rungs of the freelance jobs ladder. Developing artists who hoped to live off of their skill by selling commissions will have to compete with free AI programs. And third-world photobash monkies have another arrow in their quiver.

can you explain how this is worse than chink sweatshops that crank out mass produced art constantly?

It's cheaper, faster and there's no need to talk to chinks. I've never denied that third-world artists are a problem. It's the same principle as any other industry.

Anyone could make anything they wanted before art AI. They just had to learn art, which was a valuable, marketable skill. If you think that cheapening labor is bad for globohomo you're very confused.

Let me guess, you think things like giving everyone a platform to speak is also good for globohomo?

 

[KateMicucci]

Let me guess, you think things like giving everyone a platform to speak is also good for globohomo?

I can't tell if you're honestly trying to make a point and failing, or just doing your regular disingenuous retard act.

 

[rusty_shackleford]

Let me guess, you think things like giving everyone a platform to speak is also good for globohomo?

I can't tell if you're honestly trying to make a point and failing, or just doing your regular disingenuous retard act.

If you don't understand how letting anyone create art rather than only the people in power with the money to pay artists is bad for globohomo, I don't know what to tell you.

something something "muh human factor" as he creates another cartoon pushing trans pedo agenda on kids