Uncanny valley is the unpleasant experience one feels when confronted with something resembling a human, that isn't one (e.g., a creepy Chucky Doll). I do not know if there is a term specifically to describe the cognitive dissonance for inanimate objects.
While images like these look somewhat real you instantly get a very very unpleasant feeling, and it is very distinct from the feeling you have looking at imperfect sketches or drawings or paintings done by humans.
It's the sum of countless details that just do not fit together. Malformed objects, lighting, composition.
For lack of a better word, this image embodies the very concept of "slop". Its mere existence is truly unpleasant.
Don't get me wrong: You CAN fake reality pretty close, e.g. selloak does very very convincing impressions of real food like post #159968
the surrealist art movement was centred around objects or people that are very slightly 'off' maybe this is peaceful to someone who interprets 'reality' as 'slop'
I see your point, but on the other hand: I've been to the Dalí Theatre-Museum in Figueres a few years ago, and you really cannot compare this surrealism :)
In other words, the infamous six-finger problem we all know is evident in dozens of other ways in this image. The toilet is messed up, the toilet brush is distorted, the toilet paper, the window, the tiles, the sink, the lamps, the floor, the light, and so on and so forth...
I really don't mean to be rude, and of course I didn't downvote the image. But I think it's fair to say what I think: The model used and the task, to put it mildly, simply don't fit together. Perhaps the prompting method is just completely off.
I appreciate compliments here, but I also believe it's valuable to be honest when something simply doesn't work. (I hope I didn't express that too harshly.) Perhaps other approaches using the same or different tools or models will lead to the desired result.
Uncanny valley is the unpleasant experience one feels when confronted with something resembling a human, that isn't one (e.g., a creepy Chucky Doll). I do not know if there is a term specifically to describe the cognitive dissonance for inanimate objects.
While images like these look somewhat real you instantly get a very very unpleasant feeling, and it is very distinct from the feeling you have looking at imperfect sketches or drawings or paintings done by humans.
It's the sum of countless details that just do not fit together. Malformed objects, lighting, composition.
For lack of a better word, this image embodies the very concept of "slop". Its mere existence is truly unpleasant.
Don't get me wrong: You CAN fake reality pretty close, e.g. selloak does very very convincing impressions of real food like post #159968
I hate to see what u think of liminal horror since this is the vibe i got from this, u see AI slop, i see a dude who may be terrified of his bathroom lol
I hate to see what u think of liminal horror since this is the vibe i got from this, u see AI slop, i see a dude who may be terrified of his bathroom lol
Put this way, one could also sell people with six or seven fingers to the approvers as a "liminal horror" :) :) :)
I think the kind of horror you're describing could be created by combining the style of artwork by Vanessa Smith, Stefan Koidl, Kane Parsons, and Vilhelm Hammershøi with hotel brochures or IKEA catalogs. It would be worth a try!
Two versions, and both are censored?.. asset #642743
Yeah noticed that too lol, confusing. Now one has uncensored tits but censored pussy and the other has uncensored pussy but censored tits. And the one that is uncensored completely isn't posted.
Yeah noticed that too lol, confusing. Now one has uncensored tits but censored pussy and the other has uncensored pussy but censored tits. And the one that is uncensored completely isn't posted.
Yeah, I thought the whole thing was pretty weird, so I quickly merged the two versions together and posted in the comments in case anyone wanted it. I didn't understand why they did it that way.
Yeah noticed that too lol, confusing. Now one has uncensored tits but censored pussy and the other has uncensored pussy but censored tits. And the one that is uncensored completely isn't posted.
It was also confusing for me. But I couldn't find the uncensored one earlier neither in Twitter nor in Pixiv.
Yeah, I thought the whole thing was pretty weird, so I quickly merged the two versions together and posted in the comments in case anyone wanted it. I didn't understand why they did it that way.
Thanks for your work then! These uploaded ones is just first-party ones without third-party edits. I suppose the uploaded ones should stay "as is" as the artist uploaded them, and I can't quite do anything at this point.
Censoring pussy in Pixiv is understandable due to their policy and Japanese laws, but censoring tits in Twitter... I don't understand it too, honestly.
technically speaking, kids do not display a lot of intricate shades shown in this picture because they have a higher body fat percentage that is also where their chubby feeling comes from :)
Unfortunately, it's Lost Media; I can't find their art anymore :(, their account was deleted
Part of their generations has been preserved in the "All Uploads" section (https://aibooru.online/media_assets) — assets from #634765 down to #634570 (around page 35 at the time of writing). I think I also spotted a few more further back.
Some images still retain their metadata as well (for example, asset #634745), so the style itself isn't lost. If you have access to NAI, you can use the metadata to recreate the image and continue generating in that style. I actually saw the same style being used by someone else a few days ago, most likely a different person using the same prompt.
Also, in asset #634601, the censorship appears to have been applied incorrectly. And as far as I know, Pixiv may interpret posting large numbers of images every day as spam, which tends to attract moderator attention more quickly.
Unfortunately, it's Lost Media; I can't find their art anymore :(, their account was deleted
Ah well that is sad, not unexpected from Pixiv though site has horrible preservation. I would have a look at other sites that scrape images but it's a bit late for me, I'm sure they'll be around somewhere. Maybe a reverse image search on an image we have might bring up somewhere with the others.
konny said:
Part of their generations has been preserved in the "All Uploads" section (https://aibooru.online/media_assets) — assets from #634765 down to #634570 (around page 35 at the time of writing). I think I also spotted a few more further back.
Some images still retain their metadata as well (for example, asset #634745), so the style itself isn't lost. If you have access to NAI, you can use the metadata to recreate the image and continue generating in that style. I actually saw the same style being used by someone else a few days ago, most likely a different person using the same prompt.
Also, in asset #634601, the censorship appears to have been applied incorrectly. And as far as I know, Pixiv may interpret posting large numbers of images every day as spam, which tends to attract moderator attention more quickly.
