Schlagwörter
Aktuelle Nachrichten
America
Aus Aller Welt
Breaking News
Canada
DE
Deutsch
Deutschsprechenden
Global News
Internationale Nachrichten aus aller Welt
Japan
Japan News
Kanada
Karte
Karten
Konflikt
Korea
Krieg in der Ukraine
Latest news
Map
Maps
Nachrichten
News
News Japan
Polen
Russischer Überfall auf die Ukraine seit 2022
Science
South Korea
Ukraine
Ukraine War Video Report
UkraineWarVideoReport
United Kingdom
United States
United States of America
US
USA
USA Politics
Vereinigte Königreich Großbritannien und Nordirland
Vereinigtes Königreich
Welt
Welt-Nachrichten
Weltnachrichten
Wissenschaft
World
World News

34 Kommentare
I know AI can make more than one face so why is it always THIS face?
Eww, they are changing the artistic vision of the devs at this point. They went too far for what’s supposed to be upscaling. Make it a separate feature not called DLSS if you’re going to do that.
Related
https://www.reddit.com/r/aivideo/s/P3W31U41t3
That looks nothing like virgil
The resident evil Grace one looks awful
I think there’s a HUGE difference between upscaling+frame interpolation and whatever the hell THIS is. This isn’t „AI improving the game’s performance“ it’s adding entire details that weren’t there in the first place.
Maybe there’s a future where a game dev can make a super high resolution model and AI makes it easier to render without the need for as many polygons or something, but this is altering the whole game entirely.
Look, more wrinkles!
is dlss 4 tunred off to on the off image ? or how does it work ?
Wow so I can take that distinct AI generated look that I absolutely hate and apply it to all the characters in a game? Who thought this was a good idea?
The enshittifilter
The longer I look, the worse they get. The Resident Evil example actually impressed me a little at first until I stopped being distracted by the clarity of the image and looked more at the original. I noticed how much it changed the character’s features. It looks like it made a new face with her as a prompt, not a more realistically rendered version of the original face. The Starfield one looks weird, like the characters and background don’t exist in the same space. Matching the features seemed slightly better but still pretty off. When the Starfield one hit so weird, I went back at the first one and noticed it had the same problems.
It’s also removing things that I’m pretty sure are style choices (even if they come from system limitations). The fog is gone. The distortion from the raindrops is removed to the point of almost looking like it’s not raining. The lighting effect on characters looks like it’s from studio lights on a set, not their environment. Maybe that’s better in motion, but it really stands out to me in these pictures.
It seems like a really incredible tool being deployed „wrong“ because it feels like a good example of AI stripping the art out of something and dragging it closer to what lots of data says it „should“ look like.
Jensen: AI. AI. AI. AI. Artificial Intelligence. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. AI. Artificial Intelligence. AI. AI. AI. AI. AI. AI. AI.
Lmao why should devs and character designers bother with designing if the „upscaling“ is going to replace the finished character with an AI filter anyway?
At that point, you might as well use a beige balloon for the face.
This upscaling is stupid.
Just more AI slop, the only people that are going to be impressed by this is investors, shareholders
and gaming journalists, that’s their target audience because us gamers on the front line of the Internet have seen that AI face 100 times a day it looks nothing like Grace
It is as if Nvidia wants to get out of the consumer graphic cards to focus on AI
It literally makes everything look worse. My friend sent me this video and I can’t believe they think it looks amazing. There are some Kool-Aid drinkers out there who are more than ready to drink this down.
This is like those real time face filters that Chinese romance influencers use. This will also ruin the ability to use other graphics cards because devs will customize for this shit.
So Nvidia is essentially saying no to art direction?
Well that’s a neat idea, but it has that uncanny smoothness to it. How is this even supposed to work, it’s done locally right?
I’m not sure how I feel about this yet, please don’t dogpile me for not jumping on the bandwagon and loudly hating this lol.
Great so now developers have no idea what the game will look like on a user’s PC, and bug reports about graphics are almost useless.
Photorealistic my ass. Who told them the definition of photorealistic?
I can’t ignore the evidence of my own eyes, that is obviously an ai-ified asset and not at all photorealistic.
Take that slop away and throw it in the bin.
Real-time in game yassification, god help us
Lol, the managed to take real life scans of Virgil van dijk and apply ai to make him look like.. not Virgil van dijk
There’s something called „directions intent“ which this tech completely ignores. If I were a game developer and they were doing this to my content I’d be pretty upset.
They would do anything except digital media
It looks neat at first glance then it reminds me of those shitty AI scam advertisements for app games
Aislop 5 from slopvidia
Not only does it look horrific its being done by a second 5090 solely running dlss.
Guess that in motion the quality will tank. You can see it already with AI videos ass well.
It can only hold a consistent image for a few seconds max. You already see it with the bits shown here where clothes and background stuff is glitching out.
Outside of those issues, I just prefer to see the game like how the game designers and developers intended to and worked hard on. Even if a game has ugly models or textures, I prefer that someone worked on it. Not something a machine had made up and plagiarized.
Wow that’s atrocious. Let’s just smooth over every remaining scrap of artistry and replace every game character with the same ai slopface
Another thread full of whiners. As I’ve said elsewhere.
If you can look at the eyes on the left and say they are better than the eyes on the right, you are absolutely 100% full of shit or have terrible taste.
This is a 1:1 geometry and texture mapping with substantial improvements to lighting. Every strand of hair is 100% in place, every polygon is exactly where it is meant to be. There is absolutely nothing taking away the artistic creative vision here. This like taking a PS3 game and remastering it to a PS5 game with better shaders and the same textures.
Anyone saying this is a different face is completely wrong. It’s the same face with the same geometry and textures, with way better lighting detail. It will not impact motion quality etc. People are completely clueless. This is absolutely amazing, and anyone who doesn’t have their head 2ft up their own ass would easily acknowledge it. There is no way in hell that a blind test would favor the left image.
Imagine the artists behind the game working on this for years, and then a single piece of garbage AI comes along and ruins the character design in seconds, and it’s always the same generic filter, haha, terrifying.
Personally, I thought it looked cool. I think if they can keep it a bit truer to the original art, this could be dope!
The classic „just because you can….“ saying, etc