This preview shows page 1. Sign up to view the full content.
Unformatted text preview: sociated with xn . Nick Jones Sampling and the Brain: Inference, Control and Driving Graphic of the Graphical Model I’ll present a drawing on the board of the notional model.
I’ll suggest why this is sensible. Nick Jones Sampling and the Brain: Inference, Control and Driving Markov Random Field I
Deﬁne two Energies H1 and H2 which are minimized at zero
temperature. We seek to minimize each of the following terms
(which are coupled):
H1 = discord of perception with left eye data +
discord of perception with right eye data +
non-smoothness of perception in space +
degree of departure of perception from prior expectation +
H2 = non-smoothness of occlusion in space (left + right eye) and
amount of occlusion
I.e. a low energy choice of perception and occlusion has a smooth
perception which is a good ﬁt to both eye’s data which meets prior
expectations and is constrained by a smooth and...
View Full Document
This document was uploaded on 03/01/2014 for the course EE 208 at Imperial College.
- Spring '14