Mercurial > cortex
diff thesis/rlm-cortex-meng.tex @ 427:c670d23003de
rename.
author | Robert McIntyre <rlm@mit.edu> |
---|---|
date | Fri, 21 Mar 2014 15:36:37 -0400 |
parents | |
children | d53a31969a51 |
line wrap: on
line diff
1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000 1.2 +++ b/thesis/rlm-cortex-meng.tex Fri Mar 21 15:36:37 2014 -0400 1.3 @@ -0,0 +1,102 @@ 1.4 + 1.5 +\section{Artificial Imagination} 1.6 +\label{sec-1} 1.7 + 1.8 +Imagine watching a video of someone skateboarding. When you watch 1.9 +the video, you can imagine yourself skateboarding, and your 1.10 +knowledge of the human body and its dynamics guides your 1.11 +interpretation of the scene. For example, even if the skateboarder 1.12 +is partially occluded, you can infer the positions of his arms and 1.13 +body from your own knowledge of how your body would be positioned if 1.14 +you were skateboarding. If the skateboarder suffers an accident, you 1.15 +wince in sympathy, imagining the pain your own body would experience 1.16 +if it were in the same situation. This empathy with other people 1.17 +guides our understanding of whatever they are doing because it is a 1.18 +powerful constraint on what is probable and possible. In order to 1.19 +make use of this powerful empathy constraint, I need a system that 1.20 +can generate and make sense of sensory data from the many different 1.21 +senses that humans possess. The two key proprieties of such a system 1.22 +are \emph{embodiment} and \emph{imagination}. 1.23 + 1.24 +\subsection{What is imagination?} 1.25 +\label{sec-1-1} 1.26 + 1.27 +One kind of imagination is \emph{sympathetic} imagination: you imagine 1.28 +yourself in the position of something/someone you are 1.29 +observing. This type of imagination comes into play when you follow 1.30 +along visually when watching someone perform actions, or when you 1.31 +sympathetically grimace when someone hurts themselves. This type of 1.32 +imagination uses the constraints you have learned about your own 1.33 +body to highly constrain the possibilities in whatever you are 1.34 +seeing. It uses all your senses to including your senses of touch, 1.35 +proprioception, etc. Humans are flexible when it comes to "putting 1.36 +themselves in another's shoes," and can sympathetically understand 1.37 +not only other humans, but entities ranging from animals to cartoon 1.38 +characters to \href{http://www.youtube.com/watch?v=0jz4HcwTQmU}{single dots} on a screen! 1.39 + 1.40 + 1.41 +\begin{figure}[htb] 1.42 +\centering 1.43 +\includegraphics[width=5cm]{./images/cat-drinking.jpg} 1.44 +\caption{A cat drinking some water. Identifying this action is beyond the state of the art for computers.} 1.45 +\end{figure} 1.46 + 1.47 + 1.48 +This is a basic test for the vision system. It only tests the 1.49 +vision-pipeline and does not deal with loading eyes from a blender 1.50 +file. The code creates two videos of the same rotating cube from 1.51 +different angles. 1.52 + 1.53 + 1.54 +\begin{clojurecode} 1.55 +(in-ns 'cortex.test.vision) 1.56 + 1.57 +(defn test-pipeline 1.58 + "Testing vision: 1.59 + Tests the vision system by creating two views of the same rotating 1.60 + object from different angles and displaying both of those views in 1.61 + JFrames. 1.62 + 1.63 + You should see a rotating cube, and two windows, 1.64 + each displaying a different view of the cube." 1.65 + ([] (test-pipeline false)) 1.66 + ([record?] 1.67 + (let [candy 1.68 + (box 1 1 1 :physical? false :color ColorRGBA/Blue)] 1.69 + (world 1.70 + (doto (Node.) 1.71 + (.attachChild candy)) 1.72 + {} 1.73 + (fn [world] 1.74 + (let [cam (.clone (.getCamera world)) 1.75 + width (.getWidth cam) 1.76 + height (.getHeight cam)] 1.77 + (add-camera! world cam 1.78 + (comp 1.79 + (view-image 1.80 + (if record? 1.81 + (File. "/home/r/proj/cortex/render/vision/1"))) 1.82 + BufferedImage!)) 1.83 + (add-camera! world 1.84 + (doto (.clone cam) 1.85 + (.setLocation (Vector3f. -10 0 0)) 1.86 + (.lookAt Vector3f/ZERO Vector3f/UNIT_Y)) 1.87 + (comp 1.88 + (view-image 1.89 + (if record? 1.90 + (File. "/home/r/proj/cortex/render/vision/2"))) 1.91 + BufferedImage!)) 1.92 + (let [timer (IsoTimer. 60)] 1.93 + (.setTimer world timer) 1.94 + (display-dilated-time world timer)) 1.95 + ;; This is here to restore the main view 1.96 + ;; after the other views have completed processing 1.97 + (add-camera! world (.getCamera world) no-op))) 1.98 + (fn [world tpf] 1.99 + (.rotate candy (* tpf 0.2) 0 0)))))) 1.100 +\end{clojurecode} 1.101 + 1.102 + 1.103 +\begin{itemize} 1.104 +\item This is test1 \cite{Tappert77}. 1.105 +\end{itemize}