K. Nordberg, G. Granlund, and H. Knutsson. Representation and Learning of Invariance. In ICIP, Austin, Texas, November 1994. IEEE. Home/Search Document Details and Download
Summary Related Articles |
This paper is cited in the following contexts:
Learning in a Reactive Robotic Architecture - Andersson (2000) (Correct)
....variables only viewable via sensors percept variables. How the sensors and e ectors represent these variables is very important to know when we design our reactive architecture. We will in the following assume that the percept and response variables are represented using the channel representation [8, 13, 14, 18, 30]. In this representation we let a set of channels represent a variable. Each channel is sensitive to a small part of the variable domain and can be viewed as a band pass lter. The function we are using for the envelope of the channels is the squared cosine function. In gure 3.1 we can see the ....
....x, the di erent channels will go up and down, but using a channel function of cos 2 and an overlap of 60 degrees the sum of the channels, P c i ; and the norm of the channel vector, 30 3.2. STRUCTURE DESCRIPTION kck, will be constant. This is also true for a range of other overlaps, see [30] for a discussion about constant norm channel functions. Since y j = w T c = kwk kck cos( and kwk kck is constant, we will approximate the current response channel activity with cos( where is the angle between w and c according to the metric de ned by the channel representation. We can ....
[Article contains additional citation context not shown here]
K. Nordberg, G. Granlund, and H. Knutsson. Representation and Learning of Invariance. In ICIP, Austin, Texas, November 1994. IEEE.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC