en-es
en-fr
en-sl
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Beyond inspiration: Five lessons from biology on building intelligent machines
Published on Aug 23, 20165786 Views
The only known systems that exhibit truly intelligent, autonomous behavior are biological. If we wish to build machines that are capable of such behavior, then it makes sense to learn as much as we ca
Related categories
Chapter list
Beyond inspiration: Five lessons from biology on building intelligent machines00:00
What are the principles governing information processing in this system?01:03
Inspiration is a good start …but not enough01:06
Five lessons from biology02:52
Tiny brains/104:20
Tiny brains/204:24
Tiny brains/305:31
Jumping spider visual system09:51
Jumping spider retina10:40
Jumping spiders do object recognition12:02
Spider mimicry in flies13:03
Prey capture14:09
Navigation14:31
One-day old jumping spider/115:52
One-day old jumping spider/216:38
One-day old jumping spider/316:50
One-day old jumping spider/417:29
“Intelligence without representation19:05
Nonlinear processing in dendritic trees22:07
A brief history of neural networks 1960’s22:24
A brief history of neural networks 1980’s24:14
A brief history of neural networks 2000’s24:59
Neuron/130:48
Neuron/231:30
Bartlett Mel32:53
Sparse, overcomplete representation33:14
V1 is highly overcomplete33:27
Dense codes - Sparse, distributed codes - Local codes40:03
Sparse, distributed representation43:43
Energy function/145:31
Energy function/246:05
Coefficients ai may be computed via thresholding and lateral inhibition46:28
Examples47:49
Examples from 10x dictionary49:10
Explaining away can account for non-classical surround effects such as end-stopping/151:26
Explaining away can account for non-classical surround effects such as end-stopping/252:20
Feedback53:19
Image53:39
Hubel & Wiesel (1962, 1965)57:50
The answer58:58
Is this the goal of vision?59:38
Primate visual cortex01:00:23
Diagram01:01:13
Sensory data (x) - Actuator movement (a)01:05:33
Vision as inference01:08:36
Hierarchical Bayesian inference in visual cortex 01:09:53
What do you see?/101:11:28
What do you see?/201:12:17
Perceptual “explaining away”01:17:56
Active perception/101:18:00
Active perception/201:18:07
Human eye movements during viewing of an image01:19:00
Fixational eye movements (drift)01:19:54
Low resolution - High resolution01:21:04
Retinal ganglion cell spacing as a function of eccentricity01:22:24
Retinal ganglion cell sampling lattice (shown at one dot for every 20 ganglion cells)01:22:51
Learning the sampling lattice01:23:51
Learned glimpse window sampling lattices01:24:25
Five lessons from biology - Thank You!01:26:01