0
1
00:00:00,820 --> 00:00:07,680
In this section, first, we are going to define a neural network, talk about our brain's ability,
1

2
00:00:08,100 --> 00:00:14,400
and then we will continue the discussion by introducing a model for artificial neural networks.
2

3
00:00:14,880 --> 00:00:22,310
Let's see what can our brain do. Our brains can recognize people after so many years.
3

4
00:00:22,800 --> 00:00:26,020
This is a very important task for our brains.
4

5
00:00:26,040 --> 00:00:32,940
And if you work in the field of image processing, you might understand that this is a very complex
5

6
00:00:32,940 --> 00:00:36,120
and complicated task for a computer to do.
6

7
00:00:36,210 --> 00:00:44,100
And it requires several programming and layers to recognize a face of a human, especially if a person
7

8
00:00:44,100 --> 00:00:47,730
change the color of their hair, their color of the eyes,
8

9
00:00:47,760 --> 00:00:53,520
Computers are not really able to recognize the same person if they change a little bit.
9

10
00:00:54,000 --> 00:01:01,980
But this is a very easy task for our brain and we can recognize our best friend even after a few years,
10

11
00:01:01,980 --> 00:01:05,600
like three years, five years, without any hesitation.
11

12
00:01:06,120 --> 00:01:12,020
Next thing that our brain can do, and this is very fascinating, it's read different handwriting.
12

13
00:01:12,180 --> 00:01:16,820
The kids during the elementary school might not really be good at this task.
13

14
00:01:16,830 --> 00:01:20,220
They might not really be able to read our handwriting.
14

15
00:01:20,370 --> 00:01:27,330
But after a few years when they are in high school or better, if they are an instructor or a teacher
15

16
00:01:27,330 --> 00:01:30,780
already, they can read different handwriting.
16

17
00:01:30,900 --> 00:01:39,330
They can just read even bad handwriting. Their brain during the years of experience and seeing different
17

18
00:01:39,330 --> 00:01:41,880
handwriting improve itself.
18

19
00:01:42,070 --> 00:01:49,920
Another fascinating thing about our brain is our brain can have this flexibility and can separate the
19

20
00:01:49,920 --> 00:01:53,870
errors and false data from the useful data.
20

21
00:01:54,030 --> 00:02:01,170
So our brain has kind of filter that can filter extra lines and just read the passage itself.
21

22
00:02:01,320 --> 00:02:07,950
There are several applications for converting handwriting into a typed text.
22

23
00:02:08,230 --> 00:02:10,110
We all have it in our cell phone.
23

24
00:02:10,110 --> 00:02:15,630
You can try a Google application or other keyboards and you can try writing something.
24

25
00:02:15,870 --> 00:02:18,810
It will just translated into a text.
25

26
00:02:19,440 --> 00:02:23,790
But as you just try it for a few times, they are very sensitive.
26

27
00:02:23,790 --> 00:02:26,940
They improved during a past years very much.
27

28
00:02:26,940 --> 00:02:33,720
But, still they are sensitive and still they might get a mistake between 'a' and 'o' and some other
28

29
00:02:33,900 --> 00:02:34,570
letters.
29

30
00:02:35,040 --> 00:02:39,450
And finally, our brain can learn, which is very amazing.
30

31
00:02:39,450 --> 00:02:43,890
And that's what we are trying to implement and apply on the machines as well.
31

32
00:02:44,440 --> 00:02:52,680
A human baby, when wants to catch some object, they are actually trying, failing, trying, failing
32

33
00:02:52,680 --> 00:02:55,530
to finally attach to a specific object.
33

34
00:02:55,650 --> 00:03:02,370
Eventually, the brain is calculating the cos function based on the distance of the eyes and distance
34

35
00:03:02,370 --> 00:03:03,090
to the object.
35

36
00:03:03,360 --> 00:03:04,880
But they are not aware of that.
36

37
00:03:04,890 --> 00:03:08,670
They're just trying and failing until they can catch 
37

38
00:03:08,670 --> 00:03:10,920
the object. Our brain is amazing!
38

39
00:03:10,920 --> 00:03:15,810
Our brains can learn during the time. But let's see how can we do
39

40
00:03:15,810 --> 00:03:18,350
such an amazing and fascinating things?
40

41
00:03:18,660 --> 00:03:27,160
What is inside of our brain, of our brain has around 10 billion Neurons. Each neuron has up to ten
41

42
00:03:27,180 --> 00:03:28,710
thousand connections.
42

43
00:03:29,340 --> 00:03:36,540
Just imagine that you are living in a society that you are connected directly to ten thousand people.
43

44
00:03:37,200 --> 00:03:43,100
And whenever you have a request, whenever you're looking for something, you can directly ask the mayor,
44

45
00:03:43,100 --> 00:03:45,750
you can directly ask the president.
45

46
00:03:45,780 --> 00:03:52,110
We all know that people with too many connections, they can just easily do their so usually easier
46

47
00:03:52,110 --> 00:03:52,980
than others.
47

48
00:03:53,070 --> 00:03:55,080
Connection is very important.
48

49
00:03:55,080 --> 00:04:02,520
Even in our world! In our brain, each neuron has connections to ten thousand neurons, which is very
49

50
00:04:02,520 --> 00:04:03,180
important.
50

51
00:04:03,180 --> 00:04:08,160
And it will help of our brain to make a fast decision and do these amazing things.
51

52
00:04:08,880 --> 00:04:17,070
This is a biological model for the Neuron. A neuron takes and an input signal through the dendrites, process
52

53
00:04:17,080 --> 00:04:26,220
it likes the CPU, with Soma and process to output through a cable-like structure to other connected
53

54
00:04:26,220 --> 00:04:31,400
neurons, which is through the axon to synapses to other neurons dendrite.
54

55
00:04:31,830 --> 00:04:38,610
So this is how our brain and how the neurons inside of our brain are working.