News:

FOR INFORMATION ON DONATIONS, AND HOW TO OBTAIN ACCESS TO THE GAME, PLEASE VIEW THE FOLLOWING TOPIC: http://stick-online.com/boards/index.php?topic=2.0

Main Menu

Unlimited Detail Real-Time Rendering Technology revisit

Started by ARTgames, August 03, 2011, 10:07:03 AM

Previous topic - Next topic

Lingus

Well, they are pretty much there with faces actually. Check this out:

(Edit - Changed the link.)
http://www.emilyobrien.net/projects/image_metrics.php

Video:
http://www.youtube.com/watch?v=-r5aQYbX3bU

Implementing that into games will be a bit harder, but probably not far off. They already have games like LA Noir using similar face capture technology to capture facial expressions. It just doesn't have the photo-realistic graphics.

Jake

Quote from: Lingus on August 24, 2011, 08:18:21 PM
Well, they are pretty much there with faces actually. Check this out:

(Edit - Changed the link.)
http://www.emilyobrien.net/projects/image_metrics.php

Video:
http://www.youtube.com/watch?v=-r5aQYbX3bU

Implementing that into games will be a bit harder, but probably not far off. They already have games like LA Noir using similar face capture technology to capture facial expressions. It just doesn't have the photo-realistic graphics.
I've seen videos like this, and while I think it's nothing short of amazing, I believe we're still a long ways off (at least 10 years, and likely longer) till we can create a 3D face that our mind fully perceives as human. Immediately after I started watching the video, I picked up on subtle hints that it was a 3d model. Our brains have been conditioned to pick up on even the slightest facial gestures and cues, and anything that seems off immediately becomes grotesque and fake looking to us, even if it's 99% accurate.

Ever heard of the Uncanny Valley hypothesis?
"The uncanny valley is a hypothesis in the field of robotics[1] and 3D computer animation,[2][3] which holds that when human replicas look and act almost, but not perfectly, like actual human beings, it causes a response of revulsion among human observers. The "valley" in question is a dip in a proposed graph of the positivity of human reaction as a function of a robot's human likeness." -wikipedia

This is why in modern 3D animated movies, we often times see realistic environments with unrealistic looking people. If you try to make a realistic looking person and don't nail it perfectly, you create a disconnect with the audience, leading 3d studios to create characters with more cartoonish features, so as not to risk entering the Uncanney Valley. CGI films like Beowolf, Mars Needs Moms, and The Polar Express, were all criticized by reviewers and audience members for having "creepy" or "eerie" looking characters.

When I was watching the video you linked, I became immediately aware of issues with her smile and her eyes, and she looked almost ugly. Almost everyone that watches the video will tell you that something seems a little off, even if they can't tell you why. There is a range of theories on why this occurs, ranging from mate selection to pathogen avoidance, and that's why I think we still have a long way to go in making a 3d model truly human.

krele

Year is 2050. 'Real life' is a game with worst graphics in the world.

ARTgames

Depends what you mean by worst. I mean today you could say that's true if you don't like the look for real life. :P

Lingus

Quote from: Jake on August 25, 2011, 01:06:23 AM...I think we still have a long way to go in making a 3d model truly human.
Fair enough. Although, in my opinion, I think we're very very close. The "uncanny valley" you're talking about is, by definition, very close to reality. It's the extremely subtle differences that you're talking about that make the effect so eerie. I really think it will take very little to get past that dip and put us into the realm of photo-realistic.

ARTgames

I think we are pretty good at faking people in movies. I mean that no ware near real time 3D but are computer effects are good. I think maybe one day we will get there.