Mercurial > cortex
diff thesis/dylan-cortex-diff.diff @ 514:447c3c8405a2
accept/reject changes
author | Robert McIntyre <rlm@mit.edu> |
---|---|
date | Sun, 30 Mar 2014 10:50:05 -0400 |
parents | 4c4d45f6f30b |
children | 58fa1ffd481e |
line wrap: on
line diff
1.1 --- a/thesis/dylan-cortex-diff.diff Sun Mar 30 10:41:18 2014 -0400 1.2 +++ b/thesis/dylan-cortex-diff.diff Sun Mar 30 10:50:05 2014 -0400 1.3 @@ -294,64 +294,6 @@ 1.4 1.5 I envision =CORTEX= being used to support rapid prototyping and 1.6 iteration of ideas. Even if I could put together a well constructed 1.7 -@@ -459,8 +577,8 @@ 1.8 - simulations of very simple creatures in =CORTEX= generally run at 1.9 - 40x on my machine! 1.10 - 1.11 --** What is a sense? 1.12 -- 1.13 -+** All sense organs are two-dimensional surfaces 1.14 -+# What is a sense? 1.15 - If =CORTEX= is to support a wide variety of senses, it would help 1.16 - to have a better understanding of what a ``sense'' actually is! 1.17 - While vision, touch, and hearing all seem like they are quite 1.18 -@@ -956,7 +1074,7 @@ 1.19 - #+ATTR_LaTeX: :width 15cm 1.20 - [[./images/physical-hand.png]] 1.21 - 1.22 --** Eyes reuse standard video game components 1.23 -+** Sight reuses standard video game components... 1.24 - 1.25 - Vision is one of the most important senses for humans, so I need to 1.26 - build a simulated sense of vision for my AI. I will do this with 1.27 -@@ -1257,8 +1375,8 @@ 1.28 - community and is now (in modified form) part of a system for 1.29 - capturing in-game video to a file. 1.30 - 1.31 --** Hearing is hard; =CORTEX= does it right 1.32 -- 1.33 -+** ...but hearing must be built from scratch 1.34 -+# is hard; =CORTEX= does it right 1.35 - At the end of this section I will have simulated ears that work the 1.36 - same way as the simulated eyes in the last section. I will be able to 1.37 - place any number of ear-nodes in a blender file, and they will bind to 1.38 -@@ -1565,7 +1683,7 @@ 1.39 - jMonkeyEngine3 community and is used to record audio for demo 1.40 - videos. 1.41 - 1.42 --** Touch uses hundreds of hair-like elements 1.43 -+** Hundreds of hair-like elements provide a sense of touch 1.44 - 1.45 - Touch is critical to navigation and spatial reasoning and as such I 1.46 - need a simulated version of it to give to my AI creatures. 1.47 -@@ -2059,7 +2177,7 @@ 1.48 - #+ATTR_LaTeX: :width 15cm 1.49 - [[./images/touch-cube.png]] 1.50 - 1.51 --** Proprioception is the sense that makes everything ``real'' 1.52 -+** Proprioception provides knowledge of your own body's position 1.53 - 1.54 - Close your eyes, and touch your nose with your right index finger. 1.55 - How did you do it? You could not see your hand, and neither your 1.56 -@@ -2193,7 +2311,7 @@ 1.57 - #+ATTR_LaTeX: :width 11cm 1.58 - [[./images/proprio.png]] 1.59 - 1.60 --** Muscles are both effectors and sensors 1.61 -+** Muscles contain both sensors and effectors 1.62 - 1.63 - Surprisingly enough, terrestrial creatures only move by using 1.64 - torque applied about their joints. There's not a single straight 1.65 @@ -2440,7 +2558,8 @@ 1.66 hard control problems without worrying about physics or 1.67 senses. 1.68 @@ -371,25 +313,3 @@ 1.69 1.70 Here is the core of a basic empathy algorithm, starting with an 1.71 experience vector: 1.72 -@@ -2888,7 +3007,7 @@ 1.73 - #+end_src 1.74 - #+end_listing 1.75 - 1.76 --** Efficient action recognition with =EMPATH= 1.77 -+** =EMPATH= recognizes actions efficiently 1.78 - 1.79 - To use =EMPATH= with the worm, I first need to gather a set of 1.80 - experiences from the worm that includes the actions I want to 1.81 -@@ -3044,9 +3163,9 @@ 1.82 - to interpretation, and dissaggrement between empathy and experience 1.83 - is more excusable. 1.84 - 1.85 --** Digression: bootstrapping touch using free exploration 1.86 -- 1.87 -- In the previous section I showed how to compute actions in terms of 1.88 -+** Digression: Learn touch sensor layout through haptic experimentation, instead 1.89 -+# Boostraping touch using free exploration 1.90 -+In the previous section I showed how to compute actions in terms of 1.91 - body-centered predicates which relied averate touch activation of 1.92 - pre-defined regions of the worm's skin. What if, instead of recieving 1.93 - touch pre-grouped into the six faces of each worm segment, the true