Cosyne 2008 Workshops
March 3-4, 2008
Snow Bird, Utah
Speaker Name
H. Hill
Talk Title
Dynamic faces: behavioural studies and key questions.
Talk Abstract
Static photographs are remarkably effective stimuli given their lack of ecological validity – the first question asked behaviourally was, does motion add anything? Initial experiments using video with appropriate controls suggested motion was useful for face recognition, at least when the performance of a highly redundant system is lowered below ceiling by disrupting other cues. Subsequently, computer graphics have been used to control motion independently of shape, and to separate rigid motion of the whole head from non-rigid motion of the face. Rigid motion alone is useful for tasks including recognition and perceiving facial speech. One question is whether this distinction is reflected neurologically. The use of motion capture data to drive animations also allowed effective spatial exaggeration of facial motion, which may reflect the representation of motion. Exaggerating timing has proved more problematic, but new data shows how people do this naturally. Lastly, behavioural experiments show that movement provides an essential link between judgements of faces and voices.
