The eeriness index: quantifying the uncanny valley

An eeriness index is being developed by an Indiana University informatics professor.  Karl MacDorman is building metrics to figure out when a representation's similarity to the human is creepy, and when it's charming.  The model focuses on details like eye enlargement,  eye separation and face height.

The index also incorporates audience response.

MacDorman and his team plan to use results from objective sources like heart rate, galvanic skin response, electroencephalograms and functional magnetic imaging of the brain to gauge how far designers can go before characters creep out viewers.

 
(thanks to the twitching puppet that was once Steven Kaye)