1.1 --- a/thesis/cortex.org	Sat Mar 22 23:31:07 2014 -0400
     1.2 +++ b/thesis/cortex.org	Sun Mar 23 16:33:01 2014 -0400
     1.3 @@ -4,97 +4,52 @@
     1.4  #+description: Using embodied AI to facilitate Artificial Imagination.
     1.5  #+keywords: AI, clojure, embodiment
     1.6  
     1.7 -* Artificial Imagination
     1.8 +* Vision 
     1.9  
    1.10 -  Imagine watching a video of someone skateboarding. When you watch
    1.11 -  the video, you can imagine yourself skateboarding, and your
    1.12 -  knowledge of the human body and its dynamics guides your
    1.13 -  interpretation of the scene. For example, even if the skateboarder
    1.14 -  is partially occluded, you can infer the positions of his arms and
    1.15 -  body from your own knowledge of how your body would be positioned if
    1.16 -  you were skateboarding. If the skateboarder suffers an accident, you
    1.17 -  wince in sympathy, imagining the pain your own body would experience
    1.18 -  if it were in the same situation. This empathy with other people
    1.19 -  guides our understanding of whatever they are doing because it is a
    1.20 -  powerful constraint on what is probable and possible. In order to
    1.21 -  make use of this powerful empathy constraint, I need a system that
    1.22 -  can generate and make sense of sensory data from the many different
    1.23 -  senses that humans possess. The two key proprieties of such a system
    1.24 -  are /embodiment/ and /imagination/.
    1.25 +  System for understanding what the actors in a video are doing --
    1.26 +  Action Recognition.
    1.27 +  
    1.28 +  Separate action recognition into three components:
    1.29  
    1.30 -** What is imagination?
    1.31 +  - free play
    1.32 +  - embodied action predicates
    1.33 +  - model alignment 
    1.34 +  - sensory imagination
    1.35  
    1.36 -   One kind of imagination is /sympathetic/ imagination: you imagine
    1.37 -   yourself in the position of something/someone you are
    1.38 -   observing. This type of imagination comes into play when you follow
    1.39 -   along visually when watching someone perform actions, or when you
    1.40 -   sympathetically grimace when someone hurts themselves. This type of
    1.41 -   imagination uses the constraints you have learned about your own
    1.42 -   body to highly constrain the possibilities in whatever you are
    1.43 -   seeing. It uses all your senses to including your senses of touch,
    1.44 -   proprioception, etc. Humans are flexible when it comes to "putting
    1.45 -   themselves in another's shoes," and can sympathetically understand
    1.46 -   not only other humans, but entities ranging from animals to cartoon
    1.47 -   characters to [[http://www.youtube.com/watch?v=0jz4HcwTQmU][single dots]] on a screen!
    1.48 +* Steps 
    1.49 +  
    1.50 + - Build cortex, a simulated environment for sensate AI
    1.51 +   - solid bodies w/ joints
    1.52 +   - vision
    1.53 +   - touch
    1.54 +   - vision
    1.55 +   - hearing
    1.56 +   - proprioception
    1.57 +   - muscle contraction
    1.58  
    1.59 + - Build experimental framework for worm-actions
    1.60 +  - embodied stream predicates
    1.61 +  - \phi-space 
    1.62 +  - \phi-scan
    1.63  
    1.64 -   #+caption: A cat drinking some water. Identifying this action is beyond the state of the art for computers.
    1.65 -   #+ATTR_LaTeX: :width 5cm
    1.66 -   [[./images/cat-drinking.jpg]]
    1.67 +* News
    1.68 +  
    1.69 +  Experimental results:
    1.70  
    1.71 +  - \phi-space actually works very well for the worm!
    1.72 +  - self organizing touch map
    1.73  
    1.74 -#+begin_listing
    1.75 -\caption{This is a basic test for the vision system. It only tests the vision-pipeline and does not deal with loading eyes from a blender file. The code creates two videos of the same rotating cube from different angles.}
    1.76 -#+name: test-1
    1.77 -#+begin_src clojure
    1.78 -(defn test-pipeline
    1.79 -  "Testing vision:
    1.80 -   Tests the vision system by creating two views of the same rotating
    1.81 -   object from different angles and displaying both of those views in
    1.82 -   JFrames.
    1.83  
    1.84 -   You should see a rotating cube, and two windows,
    1.85 -   each displaying a different view of the cube."
    1.86 -  ([] (test-pipeline false))
    1.87 -  ([record?]
    1.88 -     (let [candy
    1.89 -           (box 1 1 1 :physical? false :color ColorRGBA/Blue)]
    1.90 -       (world
    1.91 -        (doto (Node.)
    1.92 -          (.attachChild candy))
    1.93 -        {}
    1.94 -        (fn [world]
    1.95 -          (let [cam (.clone (.getCamera world))
    1.96 -                width (.getWidth cam)
    1.97 -                height (.getHeight cam)]
    1.98 -            (add-camera! world cam 
    1.99 -                         (comp
   1.100 -                          (view-image
   1.101 -                           (if record?
   1.102 -                             (File. "/home/r/proj/cortex/render/vision/1")))
   1.103 -                          BufferedImage!))
   1.104 -            (add-camera! world
   1.105 -                         (doto (.clone cam)
   1.106 -                           (.setLocation (Vector3f. -10 0 0))
   1.107 -                           (.lookAt Vector3f/ZERO Vector3f/UNIT_Y))
   1.108 -                         (comp
   1.109 -                          (view-image
   1.110 -                           (if record?
   1.111 -                             (File. "/home/r/proj/cortex/render/vision/2")))
   1.112 -                          BufferedImage!))
   1.113 -            (let [timer (IsoTimer. 60)]
   1.114 -              (.setTimer world timer)
   1.115 -              (display-dilated-time world timer))
   1.116 -            ;; This is here to restore the main view
   1.117 -            ;; after the other views have completed processing
   1.118 -            (add-camera! world (.getCamera world) no-op)))
   1.119 -        (fn [world tpf]
   1.120 -          (.rotate candy (* tpf 0.2) 0 0))))))
   1.121 -#+end_src
   1.122 -#+end_listing
   1.123 +* Contributions
   1.124 +  - Built =CORTEX=, a comprehensive platform for embodied AI
   1.125 +    experiments. Has many new features lacking in other systems, such
   1.126 +    as sound. Easy to model/create new creatures.
   1.127 +  - created a novel concept for action recognition by using artificial
   1.128 +    imagination. 
   1.129  
   1.130 -- This is test1 \cite{Tappert77}.
   1.131  
   1.132 -\cite{Tappert77}
   1.133 -lol
   1.134 -\cite{Tappert77}
   1.135 \ No newline at end of file
   1.136 +
   1.137 +
   1.138 +
   1.139 +
   1.140 +