1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/thesis/abstract.tex	Sun Mar 08 22:08:17 2015 -0700
     1.3 @@ -0,0 +1,25 @@
     1.4 +Here I demonstrate the power of using embodied artificial intelligence
     1.5 +to attack the \emph{action recognition} problem, which is the challenge of
     1.6 +recognizing actions performed by a creature given limited data about
     1.7 +the creature's actions, such as a video recording. I solve this
     1.8 +problem in the case of a worm-like creature performing actions such as
     1.9 +curling and wiggling. 
    1.10 +
    1.11 +To attack the action recognition problem, I developed a computational
    1.12 +model of empathy (\texttt{EMPATH}) which allows me to recognize actions using
    1.13 +simple, embodied representations of actions (which require rich
    1.14 +sensory data), even when that sensory data is not actually available.
    1.15 +The missing sense data is imagined by combining previous experiences
    1.16 +gained from unsupervised free play. The worm is a five-segment
    1.17 +creature equipped with touch, proprioception, and muscle tension
    1.18 +senses. It recognizes actions using only proprioception data.
    1.19 +
    1.20 +In order to build this empathic, action-recognizing system, I created
    1.21 +a program called \texttt{CORTEX}, which is a complete platform for embodied
    1.22 +AI research. It provides multiple senses for simulated creatures,
    1.23 +including vision, touch, proprioception, muscle tension, and hearing.
    1.24 +Each of these senses provides a wealth of parameters that are
    1.25 +biologically inspired. \texttt{CORTEX} is able to simulate any number of
    1.26 +creatures and senses, and provides facilities for easily modeling and
    1.27 +creating new creatures. As a research platform it is more complete
    1.28 +than any other system currently available.