The pictures are nice, but the text of the "article" made me laugh out loud. A choice snippet:
Called Alter Ego, this software basically takes advantage of blend shapes to copy facial idiosyncrasies and produce emotive life-like 3D character animations.
Wow, that's so insightful! Could you copy-n-paste more of the press release, please?
Yep. In Blink, Malcolm Gladwell mentioned a couple of researchers who have identified a huge number of individual sets of facial muscles that are used in various combinations for the entire human emotional range.
I think this software -- which is incredibly amazing -- is missing just enough of those muscles that subconsciously we're getting incomplete signals from the faces.
I like the flexibility in the animations, but the mannerisms of each expression doesn't seem to fit well with the person. I wonder if this software will take advantage of characteristic expressions and not use the same animation points to produce expressions and actions.
They eyes still need lots of work. That's going to be the final frontier of the "uncanny valley". The eyes move, change focus and refract light differently hundreds of times per minute. It will make rendering realistic water look like an Amiga rendering a bouncing ball in 1989.
Comments
The pictures are nice, but the text of the "article" made me laugh out loud. A choice snippet:
Called Alter Ego, this software basically takes advantage of blend shapes to copy facial idiosyncrasies and produce emotive life-like 3D character animations.
Wow, that's so insightful! Could you copy-n-paste more of the press release, please?
Those faces are deep, DEEP in the uncanny valley, but they're not falling in, they're climbing out.
Facial animation in The Normals looks even better:
http://www.youtube.com/watch?v=Pyhf3JmODHE
Yep. In Blink, Malcolm Gladwell mentioned a couple of researchers who have identified a huge number of individual sets of facial muscles that are used in various combinations for the entire human emotional range.
I think this software -- which is incredibly amazing -- is missing just enough of those muscles that subconsciously we're getting incomplete signals from the faces.
They are not as creepy as most other animations I've seen.
I like the flexibility in the animations, but the mannerisms of each expression doesn't seem to fit well with the person. I wonder if this software will take advantage of characteristic expressions and not use the same animation points to produce expressions and actions.
They eyes still need lots of work. That's going to be the final frontier of the "uncanny valley". The eyes move, change focus and refract light differently hundreds of times per minute. It will make rendering realistic water look like an Amiga rendering a bouncing ball in 1989.
wow, I'm curious how long those took to render.
what do you think, another two decades till actors are obsolete? ;-)
"Its flexibility allows for the software to be used in real-time game cinematics, pre-rendered projects, and even game trailers."
They might be able to run in real time, though the videos are probably pre-rendered.