Download Storyboard Concept - Stanford Artificial Intelligence Laboratory

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Philosophy of artificial intelligence wikipedia , lookup

Stanford University centers and institutes wikipedia , lookup

Concept learning wikipedia , lookup

History of artificial intelligence wikipedia , lookup

Computer Go wikipedia , lookup

Convolutional neural network wikipedia , lookup

Computer vision wikipedia , lookup

Machine learning wikipedia , lookup

Ethics of artificial intelligence wikipedia , lookup

Visual servoing wikipedia , lookup

Embodied cognitive science wikipedia , lookup

Transcript
Robots and Brains
Andrew Ng, Associate Professor of Computer Science
1
Who wants a robot to clean your house?
[Photo Credit: iRobot]
2
Stanford AI Robot
[Credit: Ken Salisbury]
3
5
What’s missing?
The software
Control
6
Perception
Stanford autonomous helicopter
7
Stanford autonomous helicopter
GPS
Accelerometers
Compass
9
10
[Courtesy of David Shim]
12
Machine learning
14
Machine learning to fly helicopter
15
What’s missing?
The software
Control
16
Perception
“Robot, please find my coffee mug”
17
“Robot, please find my coffee mug”
Mug
Mug
Mug
Mug
Mug
18
Mug
Mug
Why is computer vision hard?
But the camera sees this:
19
Computer programs (features) for vision
20
SIFT
GIST
HoG
Shape context
Textons
Spin image
Why is speech recognition hard?
What a microphone records:
“Robot, please find my coffee mug.”
21
Computer programs (features) for audio
Spectrogram
Flux
22
MFCC
ZCR
Rolloff
The idea:
Most of perception in the brain
may be one simple program.
23
The “one program” hypothesis
Auditory Cortex
Auditory cortex learns to see
[Roe et al., 1992]
24
Neurons in the brain
26
Neural Network (Sparse Learning)
x1
x2
Output
x3
Layer L4
x4
Layer L1
27
Layer L3
Layer L2
How does the brain process images?
Visual cortex looks for lines/edges.
Neuron #1 of visual cortex
(model)
28
Neuron #2 of visual cortex
(model)
Comparing to Biology
Visual cortex
29
Learning algorithm
Comparing to Biology
Auditory cortex
30
Learning algorithm
Correctly found mug
33
Mistake
Correctly found mug
34
Mistake
Correctly found mug
35
Mistake
Correctly found mug
36
Mistake
Hope of progress in
Artificial Intelligence
Email: [email protected]
37
Comparing to Biology
Auditory cortex
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
Learning algorithm
50
60
20
40
40
30
50
40
60
50
50
30
30
30
30
30
50
40
40
40
40
40
60
50
50
50
50
50
60
60
60
60
60
40
60
60
80 50
60
20
20 40
40 60
60 80
2080
40
60
80
20
40
60
80
20
40
60
80
10
10
10
10
20
20
20
20
20
20
10
30
10
30
30
30
30
30
20
40
30
50
20
40
40
40
40
40
30
50
50
50
50
50
40
60
60
20
40
60
8050
20
40
60
80
60
60
20
40
20
60
40
80
60
20
40
60
80
60
10
10
20
20
30
30
40
40
50
50
60
60
80
60
20
40
60
60
2080
40
60
80
20
40
20
40
60
40
10
50
60
20
10
40
60
42
30
20
40
20
80
20
20
80
60 80
30
20
60
80
40 60
20
20
40
60
20 40
20
10
30
20
40
20
20
40
20
80
20
10
30
20
60
10
10
80
60
10
10
60
50
10
10
40
60
50
60
10
10
20
40
50
80
Missed Mugs
51
True positives
False positives
Missed Mugs
52
True positives
False positives
Missed Mugs
53
True positives
False positives
Missed Mugs
54
True positives
False positives