cortex: thesis/cortex.org comparison

comparison thesis/cortex.org @ 555:c8f7d10b1a2a

further clarification about skin partitioning per winston's request

author	Robert McIntyre <rlm@mit.edu>
date	Fri, 02 May 2014 14:27:29 -0400
parents	663e3d4f98c1
children	531bcd85d153

comparison

equal deleted inserted replaced

-:663e3d4f98c1
+:c8f7d10b1a2a
 the imagined sensory data to recognize what the observed creature is
 doing and feeling, using straightforward embodied action predicates.
 This is all demonstrated with using a simple worm-like creature, and
 recognizing worm-actions based on limited data.
-#+caption: Here is the worm with which we will be working.
+#+caption: Here is the worm with which we will be working. It is
-#+caption: It is composed of 5 segments. Each segment has a
+#+caption: composed of 5 segments. Each segment has a pair of
-#+caption: pair of extensor and flexor muscles. Each of the
+#+caption: extensor and flexor muscles. Each of the worm's four
-#+caption: worm's four joints is a hinge joint which allows
+#+caption: joints is a hinge joint which allows about 30 degrees of
-#+caption: about 30 degrees of rotation to either side. Each segment
+#+caption: rotation to either side. Each segment of the worm is
-#+caption: of the worm is touch-capable and has a uniform
+#+caption: touch-capable and has a uniform distribution of touch
-#+caption: distribution of touch sensors on each of its faces.
+#+caption: sensors on each of its faces. Each joint has a
-#+caption: Each joint has a proprioceptive sense to detect
+#+caption: proprioceptive sense to detect relative positions. The
-#+caption: relative positions. The worm segments are all the
+#+caption: worm segments are all the same except for the first one,
-#+caption: same except for the first one, which has a much
+#+caption: which has a much higher weight than the others to allow
-#+caption: higher weight than the others to allow for easy
+#+caption: for easy manual motor control.
-#+caption: manual motor control.
 #+name: basic-worm-view
 #+ATTR_LaTeX: :width 10cm
 [[./images/basic-worm-view.png]]
 #+caption: Program for reading a worm from a blender file and
 (accuracy @res)))
 #+end_src
 #+end_listing
 Running =test-empathy-accuracy= using the very short exercise
-program defined in listing \ref{generate-phi-space}, and then doing
+program =do-all-the-things= defined in listing
-a similar pattern of activity manually yields an accuracy of around
+\ref{generate-phi-space}, and then doing a similar pattern of
-73%. This is based on very limited worm experience. By training the
+activity using manual control of the worm, yields an accuracy of
-worm for longer, the accuracy dramatically improves.
+around 73%. This is based on very limited worm experience, and
+almost all errors are due to the worm's \Phi-space being too
+incomplete to properly interpret common poses. By manually training
+the worm for longer using =init-interactive= defined in listing
+\ref{manual-phi-space}, the accuracy dramatically improves:
 #+caption: Program to generate \Phi-space using manual training.
 #+name: manual-phi-space
 #+begin_listing clojure
 #+begin_src clojure
 @experiences))
 (def phi-scan (gen-phi-scan phi-space)))
 #+end_src
 #+end_listing
-After about 1 minute of manual training, I was able to achieve 95%
+=init-interactive= allows me to take direct control of the worm's
-accuracy on manual testing of the worm using =init-interactive= and
+muscles and run it through each characteristic movement I care
+about. After about 1 minute of manual training, I was able to
+achieve 95% accuracy on manual testing of the worm using
 =test-empathy-accuracy=. The majority of disagreements are near the
-transition boundaries from one type of action to another.  During
+transition boundaries from one type of action to another. During
 these transitions the exact label for the action is often unclear,
 and disagreement between empathy and experience is practically
 irrelevant. Thus, the system's effective identification accuracy is
 even higher than 95%. When I watch this system myself, I generally
 see no errors in action identification compared to my own judgment
 In the previous chapter I showed how to compute actions in terms of
 body-centered predicates, but some of those predicates relied on
 the average touch activation of pre-defined regions of the worm's
 skin. What if, instead of receiving touch pre-grouped into the six
-faces of each worm segment, the true topology of the worm's skin
+faces of each worm segment, the true partitioning of the worm's
-was unknown? This is more similar to how a nerve fiber bundle might
+skin was unknown? This is more similar to how a nerve fiber bundle
-be arranged inside an animal. While two fibers that are close in a
+might be arranged inside an animal. While two fibers that are close
-nerve bundle /might/ correspond to two touch sensors that are close
+in a nerve bundle /might/ correspond to two touch sensors that are
-together on the skin, the process of taking a complicated surface
+close together on the skin, the process of taking a complicated
-and forcing it into essentially a circle requires that some regions
+surface and forcing it into essentially a 2D circle requires that
-of skin that are close together in the animal end up far apart in
+some regions of skin that are close together in the animal end up
-the nerve bundle.
+far apart in the nerve bundle.
-In this chapter I show how to automatically learn the skin-topology of
+In this chapter I show how to automatically learn the
-a worm segment by free exploration. As the worm rolls around on the
+skin-partitioning of a worm segment by free exploration. As the
-floor, large sections of its surface get activated. If the worm has
+worm rolls around on the floor, large sections of its surface get
-stopped moving, then whatever region of skin that is touching the
+activated. If the worm has stopped moving, then whatever region of
-floor is probably an important region, and should be recorded.
+skin that is touching the floor is probably an important region,
+and should be recorded. The code I provide relies on the worm
+segment having flat faces, but still demonstrates a primitive kind
+of multi-sensory bootstrapping that I find appealing.
 #+caption: Program to detect whether the worm is in a resting state
 #+caption: with one face touching the floor.
 #+name: pure-touch
 #+begin_listing clojure
 #+begin_src clojure

Mercurial > cortex

comparison thesis/cortex.org @ 555:c8f7d10b1a2a