Testing expressions

[Manuel]

Testing the new expressions system, based on a combination of special facial actions morphs.

[Sethren]

Not bad, not bad at all.

[pistacja]

Are the ones with the closed eyes ment to have closed eyes?
'Skeptical' is great, pure mr. Spok.

I need to ask if these will be exportable?
Morphs (blend keys) are (one of) the best way to animate facial expresions and even lip-sync for a game (or some other realtime 3d)... but can I export 20-30 models to use as morph targets (or whatever the name is)? Ain't that agains the license?

[Manuel]

Are the ones with the closed eyes ment to have closed eyes?
'Skeptical' is great, pure mr. Spok.

I need to ask if these will be exportable?
Morphs (blend keys) are (one of) the best way to animate facial expresions and even lip-sync for a game (or some other realtime 3d)... but can I export 20-30 models to use as morph targets (or whatever the name is)? Ain't that agains the license?

20-30 models, if intended to be used to making an animation (and not a MH closed source clone) are ok.

Best,

Manuel

[brkurt]

Excellent work. I'm assuming that the term 'special' means you are applying Ekman's Facial Action Coding System, the international standard for
facial expressions. Check out http://www.paulekman.com for further details.

This is ground-breaking work, because one can now create micro-expressions, which are crucial for solving the
Uncanny Valley problem.

Something like this got started almost four years ago with the Geisha project, but that ran out of steam.

I'll be doing a full set of micro-expression tutorials on my website over the summer, and you've certainly sped the process up!

[Manuel]

Excellent work. I'm assuming that the term 'special' means you are applying Ekman's Facial Action Coding System, the international standard for
facial expressions. Check out http://www.paulekman.com for further details.

Not exactly. The problem of FACS is that the "sum" supposed by Ekman is not the linear sum obtained with applying 2 or more morph targets.
The "Ekman' sum" of 2 facial actions mean to "merge" the muscles that are in both the actions.
For example, take a look the result of 1 + 4: http://face-and-emotion.com/dataface/fa ... tml#914434

We are developing a different type of actions, that can be summed without problems in generic CG system.
Similar to the mpeg4 standard, but more accurate (mpeg4 is not so good for photorealistic expressions).

Best,

Manuel

[brkurt]

Outstanding! A quantitative comparison of the facial expression system you propose to Ekman's FACS, with a focus on micro-expressions,
is a worthy Ph. D. thesis. I don't need a Ph. D. at this (late) stage in my career, but I'm going to do the research anyway.

Many thanks for the boost.

[Manuel]

Outstanding! A quantitative comparison of the facial expression system you propose to Ekman's FACS, with a focus on micro-expressions,
is a worthy Ph. D. thesis. I don't need a Ph. D. at this (late) stage in my career, but I'm going to do the research anyway.

Many thanks for the boost.

Maybe you can even contribute to MH expressions development