Makehuman in the Classroom

[brkurt]

Hello everyone...

On Tuesday February 14 I'm going to be using Makehuman in my Introductory Psychology Class, partly to demonstrate Human Factors design in software, and for expression analysis using Paul Ekman's Facial Action Coding System.

I'm opening this up to forum participants: what would you like to see my students (40 between the ages of 18 and 21, 35 female and 5 male) do with Makehuman? I wouldn't mind some field-testing parameters coming from the Makehuman crew.

I'll be installing the latest Nightly Build on four Windows 7 desktops, for those interested in the technical aspects; also, I'll have older versions running on four Linux laptops.

We are going to log the entire process which I will release to the Makehuman community.

[joepal]

These are some things we've discussed at the local psychology deparment, and which we have thought about making studies concerning:

Detecting expression changes

It's possible to make an animation where a neutral face gradually slides into an expression, for example "angry". Will people with social phobia detect the target expression faster than a reference group? Will they be more or less accurate in determining which target expression is coming?

Social phobia training

We have a virtual reality studio and have discussed setting up scenes with a virtual audience so people with stage anxiety and/or social phobia can practise speaking in front of people.

...

We have a bunch of other half-baked ideas too, but the above are the most concrete. I'd love to see some experiment setups if you happen to conduct experiments with the students. (the department I'm at currently is very geared towards cognitive therapy and quantitative experimental psychology)

[brkurt]

Great idea, and easy to implement in Blender, simply by changing the Makehuman expressions every x number of frames.

I'm supposing we'd start with 24 frames as a reference then change every 12, or change only every 48.

Easy to do!

Your proposed set of experiments would be a good start, because what I'm interested in are mixed expressions that are normally part of deception.

For example: Frame 0: Neutral; Frame 12: Embarassed; Frame 24: Surprise; Frame 36: False Smile (or Smirk); Frame 48: Neutral.

So every video sequence would feature at least three different emotions: true expression; false expression; deceptive expression.

Using emoticons we would get:
Frame 0:
Frame 12:
Frame24:
Frame 36:
Frame 48:

What I would be doing is creating a strip of frames for each emoticon.