I would like to explain my matter a little bit more, because I would like to prevent any misunderstandings.

I need the expressions for the work with children with special educational needs, their teachers and parents too and for developing a special training material and (I hope so) a test for diagnosis in the near future.

workign with photos of real human faces is one solution, but one can never control, if the person at the photo will not cause any unwanted feelings in the child, that works with the card. so I want to try it with images of avatars, because they have no equivalent in the reality, but look always friendly and inspiring confidence.
but using such images, I have to ensure, that the emotional expression, that the faces of the avatars show, are as close as possible to the ones of real human faces. otherwise I cannot garantuee for the validity of my work. for example: a real human laughter cannot be faked, because the ring muscles under our eyes are changing when laughing. but they are not involved in the expression, if one is faking a laughter. because this muscles cannot be controlled by will.

so if an artist creates a beautiful laughing expression even with the finest details around the mouth region for example, with all that crinkles and so on, but makes no changes in the muscular region around the eyes, the result is (in a scientific understanding) not a real emotion. using such an image in a diagnosis test can make the test invalid.

and so I asked for expressions with a high degree of conformity with expressions to the human reality.

To say the truth, I am nit sure, if the facial structure of any of the Poser-models (figures) allows it principally to rebuild real human expressions.