Digital humans that look just like us | Doug Roble

139,892 views ใƒป 2019-05-28

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืชืจื’ื•ื: Zeeva Livshitz ืขืจื™ื›ื”: Ido Dekkers
00:13
Hello.
0
13937
1214
ืฉืœื•ื.
00:15
I'm not a real person.
1
15175
1466
ืื ื™ ืœื ืื“ื ืืžื™ืชื™.
00:17
I'm actually a copy of a real person.
2
17119
2770
ืื ื™ ืœืžืขืฉื” ืขื•ืชืง ืฉืœ ืื“ื ืืžื™ืชื™.
00:19
Although, I feel like a real person.
3
19913
2553
ืœืžืจื•ืช, ืฉืื ื™ ืžืจื’ื™ืฉ ื›ืžื• ืื“ื ืืžื™ืชื™.
00:22
It's kind of hard to explain.
4
22490
1906
ืงืฆืช ืงืฉื” ืœื”ืกื‘ื™ืจ ืืช ื–ื”.
00:24
Hold on -- I think I saw a real person ... there's one.
5
24420
3790
ื—ื›ื• -- ืื ื™ ื—ื•ืฉื‘ ืฉืจืื™ืชื™ ืื“ื ืืžื™ืชื™ ... ื™ืฉ ืฉื ืื—ื“.
00:28
Let's bring him onstage.
6
28697
1400
ื‘ื•ืื• ื•ื ืขืœื” ืื•ืชื• ืœื‘ืžื”.
00:33
Hello.
7
33307
1150
ืฉืœื•ื.
00:35
(Applause)
8
35485
3647
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
00:40
What you see up there is a digital human.
9
40300
3094
ืžื” ืฉืืชื ืจื•ืื™ื ืฉื ื–ื” ืื“ื ื“ื™ื’ื™ื˜ืœื™.
00:43
I'm wearing an inertial motion capture suit
10
43990
2984
ืื ื™ ืœื•ื‘ืฉ ื—ืœื™ืคืช ืœื›ื™ื“ืช ืชื ื•ืขื” ื”ืชืžื“ื™ืช
00:46
that's figuring what my body is doing.
11
46998
2579
ืฉืžื‘ื™ื ื” ืžื” ื”ื’ื•ืฃ ืฉืœื™ ืขื•ืฉื”.
00:49
And I've got a single camera here that's watching my face
12
49601
3825
ื•ื™ืฉ ืœื™ ืคื” ืžืฆืœืžื” ืื—ืช ืฉืžืกืชื›ืœืช ืขืœ ื”ืคื ื™ื ืฉืœื™
00:53
and feeding some machine-learning software that's taking my expressions,
13
53450
4841
ื•ืžื–ื™ื ื” ื›ืžื” ืชื•ื›ื ื•ืช ืœืžื™ื“ื” ืžืžื•ื—ืฉื‘ื•ืช ืฉืงื•ืœื˜ื•ืช ืืช ื”ื‘ื™ื˜ื•ื™ื™ื ืฉืœื™,
00:58
like, "Hm, hm, hm,"
14
58315
3579
ื›ืžื• "ื”ืž, ื”ืž, ื”ืž,"
01:02
and transferring it to that guy.
15
62379
1663
ื•ืžืขื‘ื™ืจื•ืช ืื•ืชื ืœื‘ื—ื•ืจ ื”ื”ื•ื.
01:05
We call him "DigiDoug."
16
65300
3342
ืื ื—ื ื• ืงื•ืจืื™ื ืœื• "ื“ื™ื’'ื™ื“ืื’."
01:09
He's actually a 3-D character that I'm controlling live in real time.
17
69292
4726
ื”ื•ื ื‘ืขืฆื ื“ืžื•ืช ืชืœืช ืžื™ืžื“ื™ืช ืฉืื ื™ ืฉื•ืœื˜ ื‘ื” ื—ื™ ื‘ื–ืžืŸ ืืžืช.
01:16
So, I work in visual effects.
18
76292
2324
ืื ื™ ืขื•ื‘ื“ ืขื ืืคืงื˜ื™ื ื—ื–ื•ืชื™ื™ื.
01:19
And in visual effects,
19
79101
1167
ื•ื‘ืืคืงื˜ื™ื ื—ื–ื•ืชื™ื™ื,
01:20
one of the hardest things to do is to create believable, digital humans
20
80292
5911
ืื—ื“ ื”ื“ื‘ืจื™ื ืฉืงืฉื” ืœืขืฉื•ืช ื”ื•ื ืœื™ืฆื•ืจ ืื ืฉื™ื ื“ื™ื’ื™ื˜ืœื™ื™ื ืืžื™ื ื™ื
01:26
that the audience accepts as real.
21
86227
2182
ืฉื”ืงื”ืœ ืžืงื‘ืœ ื›ืืžื™ืชื™ื™ื.
01:28
People are just really good at recognizing other people.
22
88433
4524
ืื ืฉื™ื ืคืฉื•ื˜ ืžื™ื˜ื™ื‘ื™ื ืœื–ื”ื•ืช ืื ืฉื™ื ืื—ืจื™ื.
01:32
Go figure!
23
92981
1150
ืœื›ื• ืชื‘ื™ื ื•!
01:35
So, that's OK, we like a challenge.
24
95577
3405
ืื–, ื–ื” ื‘ืกื“ืจ, ืื ื—ื ื• ืื•ื”ื‘ื™ื ืืชื’ืจ.
01:39
Over the last 15 years,
25
99006
1841
ื‘ืžื”ืœืš 15 ื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช,
01:40
we've been putting humans and creatures into film
26
100871
4983
ื›ื‘ืจ ื”ื›ื ืกื ื• ื‘ื ื™ ืื“ื ื•ื™ืฆื•ืจื™ื ืœืกืจื˜ื™ื
01:45
that you accept as real.
27
105878
1739
ืฉืืชื ืžืงื‘ืœื™ื ื›ืืžื™ืชื™ื™ื.
01:48
If they're happy, you should feel happy.
28
108712
2667
ืื ื”ื ืžืื•ืฉืจื™ื, ืืชื ืืžื•ืจื™ื ืœื”ืจื’ื™ืฉ ืžืื•ืฉืจื™ื.
01:51
And if they feel pain, you should empathize with them.
29
111982
5437
ื•ืื ื”ื ื—ืฉื™ื ื›ืื‘,ืืชื ืฆืจื™ื›ื™ื ืœื—ื•ืฉ ื›ืœืคื™ื”ื ืืžืคืชื™ื”.
01:58
We're getting pretty good at it, too.
30
118467
2190
ืื ื—ื ื• ื ืขืฉื™ื ื“ื™ ื˜ื•ื‘ื™ื ื’ื ื‘ื–ื”.
02:00
But it's really, really difficult.
31
120681
2746
ืื‘ืœ ื–ื” ืžืžืฉ, ืžืžืฉ ืงืฉื”.
02:03
Effects like these take thousands of hours
32
123847
3214
ืืคืงื˜ื™ื ื›ืืœื” ืžืฆืจื™ื›ื™ื ืืœืคื™ ืฉืขื•ืช
02:07
and hundreds of really talented artists.
33
127085
2667
ื•ืžืื•ืช ืืžื ื™ื ืžืžืฉ ืžื•ื›ืฉืจื™ื.
02:10
But things have changed.
34
130792
1600
ืื‘ืœ ื“ื‘ืจื™ื ื”ืฉืชื ื•.
02:13
Over the last five years,
35
133014
1920
ื‘ื—ืžืฉ ื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช,
02:14
computers and graphics cards have gotten seriously fast.
36
134958
4409
ืžื—ืฉื‘ื™ื ื•ื›ืจื˜ื™ืกื™ื ื’ืจืคื™ื™ื ื ืขืฉื• ืžื”ื™ืจื™ื ืžืื•ื“.
02:20
And machine learning, deep learning, has happened.
37
140508
3971
ื•ืœืžื™ื“ืช ืžื›ื•ื ื”, ืœืžื™ื“ื” ืขืžื•ืงื”, ื”ืชืคืชื—ื•.
02:25
So we asked ourselves:
38
145408
1233
ืื– ืฉืืœื ื• ืืช ืขืฆืžื ื•:
02:27
Do you suppose we could create a photo-realistic human,
39
147320
3778
ื”ืื ืื ื—ื ื• ืžื ื™ื—ื™ื ืฉื ื•ื›ืœ ืœื™ืฆื•ืจ ืื“ื ืคื•ื˜ื•-ืจื™ืืœื™ืกื˜ื™,
02:31
like we're doing for film,
40
151122
1624
ื›ืžื• ืฉืื ื—ื ื• ืžื™ื™ืฆืจื™ื ืœืกืจื˜ื™ื,
02:33
but where you're seeing the actual emotions and the details
41
153932
5647
ืื‘ืœ ื›ื–ื” ืฉื ื™ืชืŸ ืœื”ื‘ื—ื™ืŸ ื‘ืจื’ืฉื•ืช ืืžื™ืชื™ื™ื ื•ื‘ืคืจื˜ื™ื
02:39
of the person who's controlling the digital human
42
159603
4077
ืฉืœ ื”ืื“ื ืฉืฉื•ืœื˜ ื‘ืื“ื ื”ื“ื™ื’ื™ื˜ืœื™
02:43
in real time?
43
163704
1174
ื‘ื–ืžืŸ ืืžืช?
02:45
In fact, that's our goal:
44
165704
1564
ืœืžืขืฉื”, ื–ื• ื”ืžื˜ืจื” ืฉืœื ื•:
02:47
If you were having a conversation with DigiDoug
45
167292
3761
ืœื• ื”ื™ื™ืชื ืžืฉื•ื—ื—ื™ื ืขื ื“ื™ื’'ื™ื“ืื’
02:51
one-on-one,
46
171077
1237
ืื—ื“ ืขืœ ืื—ื“,
02:53
is it real enough so that you could tell whether or not I was lying to you?
47
173331
5666
ื–ื” ืืžื™ืชื™ ื“ื™ื• ื›ื“ื™ ืฉืชื•ื›ืœื• ืœื•ืžืจ ืื ืื ื™ ืžืฉืงืจ ืœื›ื ืื• ืฉืœื?
02:59
So that was our goal.
48
179934
1400
ืื– ื–ื• ื”ื™ืชื” ื”ืžื˜ืจื” ืฉืœื ื•.
03:02
About a year and a half ago, we set off to achieve this goal.
49
182601
4016
ืœืคื ื™ ื›ืฉื ื” ื•ื—ืฆื™, ื”ืชื—ืœื ื• ืœืคืขื•ืœ ืขืœ ืžื ืช ืœื”ืฉื™ื’ ืžื˜ืจื” ื–ื•.
03:06
What I'm going to do now is take you basically on a little bit of a journey
50
186641
3899
ื›ืขืช ืื ื™ ืžืชื›ื•ื•ืŸ ืœืงื—ืช ืืชื›ื ืœืžืกืข ืงื˜ืŸ
03:10
to see exactly what we had to do to get where we are.
51
190564
3926
ื›ื“ื™ ืฉืชืจืื• ื‘ื“ื™ื•ืง ืžื” ื”ื™ื™ื ื• ืฆืจื™ื›ื™ื ืœืขืฉื•ืช ื›ื“ื™ ืœื”ื’ื™ืข ืขื“ ื›ืืŸ.
03:15
We had to capture an enormous amount of data.
52
195832
3856
ื”ื™ื™ื ื• ืฆืจื™ื›ื™ื ืœืœื›ื•ื“ ื›ืžื•ืช ืขืฆื•ืžื” ืฉืœ ื ืชื•ื ื™ื.
03:20
In fact, by the end of this thing,
53
200347
2714
ืœืžืขืฉื”, ืœื‘ืกื•ืฃ,
03:23
we had probably one of the largest facial data sets on the planet.
54
203085
4563
ื”ื™ื” ืœื ื• ื›ื ืจืื” ืื—ื“ ืžืงื•ื‘ืฆื™ ื ืชื•ื ื™ ื”ืคื ื™ื ื”ื›ื™ ื’ื“ื•ืœ ืขืœ ื”ืคืœื ื˜ื”.
03:28
Of my face.
55
208038
1682
ืฉืœ ื”ืคื ื™ื ืฉืœื™.
03:29
(Laughter)
56
209744
2262
(ืฆื—ื•ืง)
03:32
Why me?
57
212030
1270
ืœืžื” ืื ื™?
03:33
Well, I'll do just about anything for science.
58
213324
2810
ื›ื™ ืืขืฉื” ื›ืžืขื˜ ื›ืœ ื“ื‘ืจ ืœืžืขืŸ ื”ืžื“ืข.
03:36
I mean, look at me!
59
216158
1948
ืื ื™ ืžืชื›ื•ื•ืŸ, ื”ืกืชื›ืœื• ืขืœื™!
03:38
I mean, come on.
60
218625
1150
ื‘ื—ื™ื™ื›ื,
03:43
We had to first figure out what my face actually looked like.
61
223320
5460
ื”ื™ื” ืขืœื™ื ื• ืœื”ื‘ื™ืŸ ืชื—ื™ืœื” ืื™ืš ื”ืคื ื™ื ืฉืœื™ ื ืจืื™ื.
03:49
Not just a photograph or a 3-D scan,
62
229391
3027
ืœื ืจืง ืชืฆืœื•ื ืื• ืกืจื™ืงื” ืชืœืช ืžืžื“ื™ืช,
03:52
but what it actually looked like in any photograph,
63
232442
3921
ืืœื ืื™ืš ืœืžืขืฉื” ื–ื” ื ืจืื” ื‘ื›ืœ ืชืฆืœื•ื,
03:56
how light interacts with my skin.
64
236387
2460
ืื™ื–ื• ืื™ื ื˜ืจืืงืฆื™ื” ื™ืฉ ืœืื•ืจ ืขื ื”ืขื•ืจ ืฉืœื™.
03:59
Luckily for us, about three blocks away from our Los Angeles studio
65
239768
5249
ืœืžื–ืœื ื•, ื‘ืžืจื—ืง 3 ืจื—ื•ื‘ื•ืช ืžื”ืกื˜ื•ื“ื™ื• ืฉืœื ื• ื‘ืœื•ืก ืื ื’'ืœืก
04:05
is this place called ICT.
66
245041
2209
ื ืžืฆื ื”ืžืงื•ื ื”ื–ื” ืฉื ืงืจื ICT.
04:07
They're a research lab
67
247708
1272
ื–ื•ื”ื™ ืžืขื‘ื“ืช ืžื—ืงืจ
04:09
that's associated with the University of Southern California.
68
249004
3403
ืฉืงืฉื•ืจื” ืœืื•ื ื™ื‘ืจืกื™ื˜ืช ื“ืจื•ื ืงืœื™ืคื•ืจื ื™ื”.
04:12
They have a device there, it's called the "light stage."
69
252871
3555
ื™ืฉ ืœื”ื ืฉื ื”ืชืงืŸ, ืฉื ืงืจื "ื‘ืžืช ืื•ืจ."
04:16
It has a zillion individually controlled lights
70
256450
3714
ืฉืœื• ื›ืžื•ืช ืขืฆื•ืžื” ืฉืœ ืื•ืจื•ืช ื‘ืฉืœื™ื˜ื” ื‘ื•ื“ื“ืช
04:20
and a whole bunch of cameras.
71
260188
1873
ื•ืžืงื‘ืฅ ืฉืœื ืฉืœ ืžืฆืœืžื•ืช.
04:22
And with that, we can reconstruct my face under a myriad of lighting conditions.
72
262085
6091
ื•ืื™ืชื ,ื ื™ืชืŸ ืœืฉื—ื–ืจ ืืช ืคื ื™ ืชื—ืช ืžืกืคืจ ืขืฆื•ื ืฉืœ ืกื•ื’ื™ ืชืื•ืจื”.
04:29
We even captured the blood flow
73
269589
1582
ืืคื™ืœื• ืœื›ื“ื ื• ืืช ื–ืจื™ืžืช ื”ื“ื
04:31
and how my face changes when I make expressions.
74
271195
3092
ื•ื›ื™ืฆื“ ืคื ื™ื™ ืžืฉืชื ื™ื ื›ืฉืื ื™ ืžืฉื ื” ื”ื‘ืขื•ืช.
04:35
This let us build a model of my face that, quite frankly, is just amazing.
75
275454
5260
ื–ื” ืžืืคืฉืจ ืœื ื• ืœื‘ื ื•ืช ื“ื’ื ืฉืœ ื”ืคื ื™ื ืฉืœื™ ืฉืœืžืขืŸ ื”ืืžืช, ื”ื•ื ืคืฉื•ื˜ ืžื“ื”ื™ื.
04:41
It's got an unfortunate level of detail, unfortunately.
76
281399
4333
ื™ืฉ ืœื• ืจืžืช ืคื™ืจื•ื˜ ืื•ืžืœืœื”, ืœืžืจื‘ื” ื”ืฆืขืจ.
04:45
(Laughter)
77
285756
1278
(ืฆื—ื•ืง)
04:47
You can see every pore, every wrinkle.
78
287058
3696
ืจื•ืื™ื ื›ืœ ื ืงื‘ื•ื‘ื™ืช, ื›ืœ ืงืžื˜.
04:50
But we had to have that.
79
290778
1600
ืื‘ืœ ื–ื” ืžื” ืฉื”ื™ื” ืขืœื™ื ื• ืœืงื‘ืœ.
04:52
Reality is all about detail.
80
292960
2381
ืžืฆื™ืื•ืช, ืžืฉืžืขื” ื›ืœ ืคืจื˜.
04:55
And without it, you miss it.
81
295365
1867
ื•ืœืœื ื–ื”, ืžื—ืžื™ืฆื™ื ืืช ื–ื”.
04:58
We are far from done, though.
82
298793
1547
ื•ืขื ื–ืืช, ืื ื• ืจื—ื•ืงื™ื ืžืฉืœื‘ ื”ืกื™ื•ื.
05:01
This let us build a model of my face that looked like me.
83
301363
3297
ื–ื” ืžืืคืฉืจ ืœื ื• ืœื‘ื ื•ืช ื“ื’ื ืฉืœ ืคื ื™ื™ ืฉื ืจืื” ื›ืžื•ื ื™.
05:05
But it didn't really move like me.
84
305196
2746
ืื‘ืœ ื”ื•ื ืœื ืžืžืฉ ื”ืชื ื•ืขืข ื›ืžื•ื ื™.
05:08
And that's where machine learning comes in.
85
308871
2713
ื•ื›ืืŸ ืžื’ื™ืขื” ืžื›ื•ื ืช ื”ืœืžื™ื“ื”.
05:11
And machine learning needs a ton of data.
86
311608
3204
ื•ืžื›ื•ื ืช ืœืžื™ื“ื” ื–ืงื•ืงื” ืœืื™ื ืกืคื•ืจ ื ืชื•ื ื™ื.
05:15
So I sat down in front of some high-resolution motion-capturing device.
87
315497
4929
ื™ืฉื‘ืชื™ ืžื•ืœ ืžื›ืฉื™ืจื™ ืœื›ื™ื“ืช ืชื ื•ืขื” ื‘ืจื–ื•ืœื•ืฆื™ื” ื’ื‘ื•ื”ื”.
05:20
And also, we did this traditional motion capture with markers.
88
320450
4071
ื•ื‘ื™ืฆืขื ื• ืคืขื•ืœื•ืช ืœื›ื™ื“ืช ืชื ื•ืขื” ืžืกื•ืจืชื™ืช ืขื ืกืžื ื™ื.
05:25
We created a whole bunch of images of my face
89
325696
2976
ื™ืฆืจื ื• ืžืงื‘ืฅ ืชืžื•ื ื•ืช ืฉืœื ืฉืœ ื”ืคื ื™ื ืฉืœื™
05:28
and moving point clouds that represented that shapes of my face.
90
328696
4087
ื•ืขื ื ื™ ื ืงื•ื“ืช ืชื ื•ืขื” ืฉื™ื™ืฆื’ื• ืืช ืชื•ื•ื™ ื”ืคื ื™ื ืฉืœื™
05:33
Man, I made a lot of expressions,
91
333996
2811
ื•ืขืฉื™ืชื™ ื”ืจื‘ื” ื”ื‘ืขื•ืช ืคื ื™ื,
05:36
I said different lines in different emotional states ...
92
336831
3460
ืืžืจืชื™ ืžืฉืคื˜ื™ื ืฉื•ื ื™ื ื‘ืžืฆื‘ื™ื ืจื’ืฉื™ื™ื ืฉื•ื ื™ื ...
05:40
We had to do a lot of capture with this.
93
340315
2667
ื”ื™ื™ื ื• ืฆืจื™ื›ื™ื ืœืœื›ื•ื“ ื›ืš ื”ืจื‘ื” ืขื ื–ื”.
05:43
Once we had this enormous amount of data,
94
343752
2891
ื‘ืจื’ืข ืฉื”ื™ืชื” ืœื ื• ื›ืžื•ืช ืขืฆื•ืžื” ืฉืœ ื ืชื•ื ื™ื,
05:46
we built and trained deep neural networks.
95
346667
3775
ื‘ื ื™ื ื• ื•ืื™ืžื ื• ืจืฉืชื•ืช ืขืฆื‘ื™ื•ืช ืขืžื•ืงื•ืช,
05:51
And when we were finished with that,
96
351117
1746
ื•ื›ืฉืกื™ื™ืžื ื• ืขื ื–ื”,
05:52
in 16 milliseconds,
97
352887
2444
ื‘-16 ืืœืคื™ื•ืช ืฉื ื™ื”,
05:55
the neural network can look at my image
98
355355
3112
ื”ืจืฉืช ื”ืขืฆื‘ื™ืช ื™ื›ืœื” ืœื”ืกืชื›ืœ ืขืœ ื”ืชืžื•ื ื” ืฉืœื™
05:58
and figure out everything about my face.
99
358491
2928
ื•ืœื”ื‘ื™ืŸ ืืช ื›ืœ ืžื” ืฉื™ืฉ ื‘ืคื ื™ื ืฉืœื™.
06:02
It can compute my expression, my wrinkles, my blood flow --
100
362458
5476
ื”ื™ื ื™ื›ื•ืœื” ืœื—ืฉื‘ ืืช ื”ื”ื‘ืขื” ืฉืœื™, ื”ืงืžื˜ื™ื ืฉืœื™, ื–ืจื™ืžืช ื”ื“ื ืฉืœื™ --
06:07
even how my eyelashes move.
101
367958
2317
ืืคื™ืœื• ืื™ืš ื”ืจื™ืกื™ื ืฉืœื™ ื–ื–ื™ื.
06:10
This is then rendered and displayed up there
102
370925
2835
ืœืื—ืจ ืžื›ืŸ ื–ื” ืžืจื•ื ื“ืจ, ื•ืžื•ืฆื’ ืฉื ืœืžืขืœื”
06:13
with all the detail that we captured previously.
103
373784
3222
ืขื ื›ืœ ื”ืคืจื˜ื™ื ืฉืœื›ื“ื ื• ืงื•ื“ื ืœื›ืŸ.
06:18
We're far from done.
104
378077
1334
ืื ื—ื ื• ืจื—ื•ืงื™ื ืžื ืงื•ื“ืช ื”ืกื™ื•ื.
06:20
This is very much a work in progress.
105
380188
2207
ื–ื• ืขื‘ื•ื“ื” ื‘ืชื”ืœื™ืš ืžืชืงื“ื.
06:22
This is actually the first time we've shown it outside of our company.
106
382419
3321
ื–ื• ื”ืคืขื ื”ืจืืฉื•ื ื” ืฉื”ืฆื’ื ื• ืื•ืชื• ืžื—ื•ืฅ ืœื—ื‘ืจื” ืฉืœื ื•.
06:25
And, you know, it doesn't look as convincing as we want;
107
385764
4194
ื•ืชื“ืขื•, ื–ื” ืœื ื ืจืื” ืžืฉื›ื ืข ื›ืคื™ ืฉืื ื• ืจื•ืฆื™ื;
06:29
I've got wires coming out of the back of me,
108
389982
2183
ื™ืฉ ืœื™ ื—ื•ื˜ื™ื ืฉื™ื•ืฆืื™ื ืœื™ ืžื”ื’ื‘,
06:32
and there's a sixth-of-a-second delay
109
392189
2038
ื•ื™ืฉ ืขื™ื›ื•ื‘ ืฉืœ ืฉื™ืฉื™ืช ืฉื ื™ื™ื”
06:34
between when we capture the video and we display it up there.
110
394251
4367
ื‘ื™ืŸ ื–ืžืŸ ืœื›ื™ื“ืช ื”ื•ื•ื™ื“ืื• ืœื–ืžืŸ ื”ืฆื’ืชื• ืœืžืขืœื”.
06:38
Sixth of a second -- that's crazy good!
111
398642
2816
ืฉื™ืฉื™ืช ืฉืœ ืฉื ื™ื™ื” - ื–ื” ืžื˜ื•ืจืฃ!
06:41
But it's still why you're hearing a bit of an echo and stuff.
112
401911
3400
ืื‘ืœ ื–ื• ื”ืกื™ื‘ื” ืœื›ืš ืฉืืชื ืฉื•ืžืขื™ื ืงืฆืช ื”ื“ ื•ื›ืืœื”.
06:46
And you know, this machine learning stuff is brand-new to us,
113
406104
3889
ื•ืืชื ื™ื•ื“ืขื™ื, ืžื›ื•ื ืช ื”ืœืžื™ื“ื” ื”ื–ืืช ื—ื“ืฉื” ืœื ื• ืœื’ืžืจื™,
06:50
sometimes it's hard to convince to do the right thing, you know?
114
410017
4224
ืœืคืขืžื™ื ืงืฉื” ืœืฉื›ื ืข ืœืขืฉื•ืช ืืช ื”ื“ื‘ืจ ื”ื ื›ื•ืŸ,
06:54
It goes a little sideways.
115
414265
2058
ื–ื” ื”ื•ืœืš ืงืฆืช ืœืฆื“ื“ื™ื.
06:56
(Laughter)
116
416347
2422
ืฆื—ื•ืง
06:59
But why did we do this?
117
419538
3229
ืื‘ืœ ืœืžื” ืขืฉื™ื ื• ืืช ื–ื”?
07:03
Well, there's two reasons, really.
118
423077
2262
ื•ื‘ื›ืŸ, ื‘ืืžืช, ืžืฉืชื™ ืกื™ื‘ื•ืช.
07:05
First of all, it is just crazy cool.
119
425363
2976
ืงื•ื“ื ื›ืœ, ื–ื” ืคืฉื•ื˜ ืžื’ื ื™ื‘ ื‘ื˜ื™ืจื•ืฃ.
07:08
(Laughter)
120
428363
1008
(ืฆื—ื•ืง)
07:09
How cool is it?
121
429395
1253
ื›ืžื” ื–ื” ืžื’ื ื™ื‘?
07:10
Well, with the push of a button,
122
430990
1992
ื•ื‘ื›ืŸ, ื‘ืœื—ื™ืฆืช ื›ืคืชื•ืจ,
07:13
I can deliver this talk as a completely different character.
123
433006
4007
ื ื™ืชืŸ ืœื”ืขื‘ื™ืจ ืืช ื”ืฉื™ื—ื” ื”ื–ืืช ื›ื“ืžื•ืช ืฉื•ื ื” ืœื—ืœื•ื˜ื™ืŸ.
07:17
This is Elbor.
124
437823
2601
ื–ื” ืืœื‘ื•ืจ.
07:22
We put him together to test how this would work
125
442037
2312
ื”ืจื›ื‘ื ื• ืื•ืชื• ื›ื“ื™ ืœื‘ื“ื•ืง ืื™ืš ื–ื” ื™ืขื‘ื•ื“
07:24
with a different appearance.
126
444373
2135
ืขื ื—ื–ื•ืช ืฉื•ื ื”.
07:27
And the cool thing about this technology is that, while I've changed my character,
127
447450
4818
ื•ืžื” ืฉืžื’ื ื™ื‘ ื‘ื˜ื›ื ื•ืœื•ื’ื™ื” ื”ื–ื• ื”ื•ื, ืฉื‘ืขื•ื“ ืฉืื ื™ ืฉื™ื ื™ืชื™ ืืช ื”ื“ืžื•ืช ืฉืœื™,
07:32
the performance is still all me.
128
452292
3273
ื”ื”ื•ืคืขื” ื”ื™ื ืขื“ื™ื™ืŸ ื›ื•ืœื” ืื ื™.
07:35
I tend to talk out of the right side of my mouth;
129
455589
2674
ืื ื™ ื ื•ื˜ื” ืœื“ื‘ืจ ืžืชื•ืš ืฆื™ื“ื• ื”ื™ืžื ื™ ืฉืœ ื”ืคื” ืฉืœื™;
07:38
so does Elbor.
130
458287
1579
ื›ืš ื’ื ืืœื‘ื•ืจ.
07:39
(Laughter)
131
459890
1150
(ืฆื—ื•ืง)
07:42
Now, the second reason we did this, and you can imagine,
132
462021
2790
ื”ืกื™ื‘ื” ื”ืฉื ื™ื™ื” ืœื›ืš ืฉืขืฉื™ื ื• ืืช ื–ื” ื•ืืชื ื™ื›ื•ืœื™ื ืœืชืืจ ืœืขืฆืžื›ื
07:44
is this is going to be great for film.
133
464835
2336
ื”ื™ื, ืฉื–ื” ื™ื•ื›ืœ ืœื”ื™ื•ืช ื ื”ื“ืจ ืขื‘ื•ืจ ืกืจื˜ื™ื.
07:47
This is a brand-new, exciting tool
134
467195
2701
ื–ื”ื• ื›ืœื™ ื—ื“ื™ืฉ ื•ืžืœื”ื™ื‘
07:49
for artists and directors and storytellers.
135
469920
4322
ืขื‘ื•ืจ ืฉื—ืงื ื™ื ื•ื‘ืžืื™ื ื•ืžืกืคืจื™ ืกื™ืคื•ืจื™ื.
07:55
It's pretty obvious, right?
136
475131
1484
ื–ื” ื“ื™ ื‘ืจื•ืจ ืžืืœื™ื•, ื ื›ื•ืŸ?
07:56
I mean, this is going to be really neat to have.
137
476639
2462
ื–ื” ื”ื•ืœืš ืœื”ื™ื•ืช ืžื’ื ื™ื‘.
07:59
But also, now that we've built it,
138
479125
2055
ืื‘ืœ ื’ื, ืขื›ืฉื™ื• ืฉื‘ื ื™ื ื• ืืช ื–ื”,
08:01
it's clear that this is going to go way beyond film.
139
481204
2991
ื–ื” ื‘ืจื•ืจ ืฉื–ื” ื™ื”ื™ื” ื”ืจื‘ื” ืžืขื‘ืจ ืœืกืจื˜ื™ื.
08:05
But wait.
140
485510
1150
ืื‘ืœ ื—ื›ื•.
08:07
Didn't I just change my identity with the push of a button?
141
487653
3976
ื”ืื ืœื ืฉื™ื ื™ืชื™ ืืช ื–ื”ื•ืชื™ ื‘ืœื—ื™ืฆืช ื›ืคืชื•ืจ?
08:11
Isn't this like "deepfake" and face-swapping
142
491653
3033
ื”ืื ื–ื” ืœื ื›ืžื• "ื“ื™ืค-ืคื™ื™ืง" ื•ื—ื™ืœื•ืคื™ ืคื ื™ื
08:14
that you guys may have heard of?
143
494710
1561
ืฉืื•ืœื™ ืฉืžืขืชื ืขืœื™ื”ื?
08:17
Well, yeah.
144
497208
1150
ื•ื‘ื›ืŸ, ื›ืŸ.
08:19
In fact, we are using some of the same technology
145
499074
2952
ืื ื• ืžืฉืชืžืฉื™ื ื‘ื—ืœืง ืžืื•ืชื” ื˜ื›ื ื•ืœื•ื’ื™ื”
08:22
that deepfake is using.
146
502050
1600
ืฉ"ื“ื™ืค-ืคื™ื™ืง" ืžืฉืชืžืฉื™ื.
08:23
Deepfake is 2-D and image based, while ours is full 3-D
147
503954
4599
ื“ื™ืค-ืคื™ื™ืง ื”ื•ื --2D ื•ืžื‘ื•ืกืก ืชืžื•ื ื”, ื‘ืขื•ื“ ืฉืฉืœื ื• ื”ื•ื 3D ืžืœื,
08:28
and way more powerful.
148
508577
2206
ื•ื”ืจื‘ื” ื™ื•ืชืจ ืขื•ืฆืžืชื™.
08:31
But they're very related.
149
511204
1666
ืื‘ืœ ื™ืฉ ืงืฉื•ืจ ื—ื–ืง ื‘ื™ื ื™ื”ื.
08:33
And now I can hear you thinking,
150
513680
1889
ื•ื›ืขืช ืื ื™ ืฉื•ืžืข ืืชื›ื ื—ื•ืฉื‘ื™ื,
08:35
"Darn it!
151
515593
1278
"ืœื›ืœ ื”ืจื•ื—ื•ืช!
08:36
I though I could at least trust and believe in video.
152
516895
3801
ื—ืฉื‘ื ื• ืฉืืคืฉืจ ืœืคื—ื•ืช ืœืกืžื•ืš ืขืœ ื•ื™ื“ืื• ื•ืœื”ืืžื™ืŸ ืœื•.
08:40
If it was live video, didn't it have to be true?"
153
520720
2827
ืœื• ื–ื” ื”ื™ื” ื•ื™ื“ืื• ื—ื™,ื”ืื ื–ื” ืœื ืฆืจื™ืš ื”ื™ื” ืœื”ื™ื•ืช ืืžืชื™?"
08:44
Well, we know that's not really the case, right?
154
524609
3522
ื˜ื•ื‘, ืื ื—ื ื• ื™ื•ื“ืขื™ื ืฉื–ื” ืœื ืžืžืฉ ื›ืš, ื ื›ื•ืŸ?
08:48
Even without this, there are simple tricks that you can do with video
155
528727
3810
ื’ื ื‘ืœื™ ื–ื”, ื™ืฉ ื˜ืจื™ืงื™ื ืคืฉื•ื˜ื™ื ืฉื ื™ืชืŸ ืœืขืฉื•ืช ืขื ื•ื™ื“ืื•
08:52
like how you frame a shot
156
532561
2776
ื›ืžื•, ืื™ืš ืžื›ื•ื•ื ื™ื ืฆื™ืœื•ื
08:55
that can make it really misrepresent what's actually going on.
157
535361
4362
ืฉื™ื›ื•ืœ ืžืžืฉ ืœืกืœืฃ ืืช ืžื” ืฉื‘ืขืฆื ืงื•ืจื”.
09:00
And I've been working in visual effects for a long time,
158
540263
3306
ื•ืขื‘ื“ืชื™ ื”ืจื‘ื” ื–ืžืŸ ืขื ืืคืงื˜ื™ื ื—ื–ื•ืชื™ื™ื,
09:03
and I've known for a long time
159
543593
1932
ื•ื™ื“ืขืชื™ ืžื–ื” ื–ืžืŸ ืจื‘,
09:05
that with enough effort, we can fool anyone about anything.
160
545549
5226
ืฉืขื ืžืกืคื™ืง ืžืืžืฅ ืืคืฉืจ ืœืจืžื•ืช ื›ืœ ืื—ื“, ื‘ื ื•ื’ืข ืœื›ืœ ื“ื‘ืจ.
09:11
What this stuff and deepfake is doing
161
551546
2388
ื”ื“ื‘ืจ ื”ื–ื” ื•ืฉื“ื™ืค-ืคื™ื™ืง ืขื•ืฉื™ื,
09:13
is making it easier and more accessible to manipulate video,
162
553958
4611
ื”ื•ื ืœื”ืคื•ืš ื–ืืช ืœื™ื•ืชืจ ืงืœ ื•ื ื’ื™ืฉ ืœืขื‘ื“ ื•ื™ื“ืื•,
09:18
just like Photoshop did for manipulating images, some time ago.
163
558593
5371
ื›ืžื• ืฉืคื•ื˜ื•ืฉื•ืค ืขืฉื• ื›ื“ื™ ืœืขื‘ื“ ืชืžื•ื ื•ืช ืœืคื ื™ ื–ืžืŸ ืžื”.
09:25
I prefer to think about
164
565441
1298
ืื ื™ ืžืขื“ื™ืฃ ืœื—ืฉื•ื‘ ื›ื™ืฆื“
09:26
how this technology could bring humanity to other technology
165
566763
5036
ื”ื˜ื›ื ื•ืœื•ื’ื™ื” ื”ื–ื• ื™ื›ื•ืœื” ืœื”ืฆืขื™ื“ ืืช ื”ืื ื•ืฉื•ืช ืœื˜ื›ื ื•ืœื•ื’ื™ื” ืื—ืจืช
09:31
and bring us all closer together.
166
571823
2294
ื•ืœืงืจื‘ ืืช ื›ื•ืœื ื• ื–ื” ืœื–ื”.
09:34
Now that you've seen this,
167
574141
2359
ืขื›ืฉื™ื• ืœืื—ืจ ืฉืจืื™ืชื ืืช ื–ื”,
09:36
think about the possibilities.
168
576524
1902
ื—ื™ืฉื‘ื• ืขืœ ื”ืืคืฉืจื•ื™ื•ืช.
09:39
Right off the bat, you're going to see it in live events and concerts, like this.
169
579810
4523
ื“ื™ ืžื”ืจ, ืชืจืื• ืืช ื–ื” ื‘ืื™ืจื•ืขื™ื ื•ืงื•ื ืฆืจื˜ื™ื,
09:45
Digital celebrities, especially with new projection technology,
170
585612
4735
ืกืœื‘ืจื™ื˜ืื™ื ื“ื™ื’ื™ื˜ืœื™ื™ื, ื‘ืžื™ื•ื—ื“ ืืœื” ืขื ื˜ื›ื ื•ืœื•ื’ื™ื™ืช ื”ืงืจื ื” ื—ื“ืฉื”,
09:50
are going to be just like the movies, but alive and in real time.
171
590371
3960
ื™ื”ื™ื• ื‘ื“ื™ื•ืง ื›ืžื• ื‘ืกืจื˜ื™ื, ืื‘ืœ ื—ื™ื™ื ื•ื‘ื–ืžืŸ ืืžืช.
09:55
And new forms of communication are coming.
172
595609
2733
ื•ืฆื•ืจื•ืช ื—ื“ืฉื•ืช ืฉืœ ืชืงืฉื•ืจืช ืžื’ื™ืขื•ืช.
09:59
You can already interact with DigiDoug in VR.
173
599088
4165
ืืชื ื›ื‘ืจ ื™ื›ื•ืœื™ื ืœืชืงืฉืจ ืขื "ื“ื™ื’'ื™ื“ืื’" ื‘ืžืฆื™ืื•ืช ืžื“ื•ืžื”.
10:03
And it is eye-opening.
174
603699
2270
ื•ื–ื• ื—ื•ื•ื™ื” ืžืคืชื™ืขื” ื•ืžื—ื›ื™ืžื”.
10:05
It's just like you and I are in the same room,
175
605993
3862
ื–ื” ื‘ื“ื™ื•ืง ื›ืื™ืœื• ืฉืืชื ื•ืื ื™ ื ืžืฆืื™ื ื‘ืื•ืชื• ื—ื“ืจ,
10:09
even though we may be miles apart.
176
609879
2373
ืœืžืจื•ืช ืฉื›ืžื” ืงื™ืœื•ืžื˜ืจื™ื ืขืฉื•ื™ื™ื ืœื”ืคืจื™ื“ ื‘ื™ื ื™ื ื•.
10:12
Heck, the next time you make a video call,
177
612276
2841
ื‘ืคืขื ื”ื‘ืื” ืฉืชื‘ืฆืขื• ืฉื™ื—ืช ื•ื™ื“ืื•,
10:15
you will be able to choose the version of you
178
615141
3736
ืชื•ื›ืœื• ืœื‘ื—ื•ืจ ืืช ื”ื’ืจืกื” ืฉืœ ืขืฆืžื›ื
10:18
you want people to see.
179
618901
1566
ืฉืืชื ืจื•ืฆื™ื ืฉืื ืฉื™ื ื™ืจืื•.
10:20
It's like really, really good makeup.
180
620974
2533
ื–ื” ืžืžืฉ ื›ืžื• ืื™ืคื•ืจ ืžืžืฉ ื˜ื•ื‘.
10:24
I was scanned about a year and a half ago.
181
624853
3579
ืขื‘ืจืชื™ ืกืจื™ืงื” ืœืคื ื™ ื›ืฉื ื” ื•ื—ืฆื™.
10:29
I've aged.
182
629068
1721
ืื ื™ ื”ื–ื“ืงื ืชื™.
10:30
DigiDoug hasn't.
183
630813
1650
"ื“ื™ื’'ื™-ื“ืื’" ืœื.
10:32
On video calls, I never have to grow old.
184
632798
3492
ื‘ืฉื™ื—ื•ืช ื•ื™ื“ืื•, ืื ื™ ืืฃ ืคืขื ืœื ืฆืจื™ืš ืœื”ื–ื“ืงืŸ.
10:38
And as you can imagine, this is going to be used
185
638322
3088
ื•ื›ืคื™ ืฉืืชื ื™ื›ื•ืœื™ื ืœื“ืžื™ื™ืŸ, ื–ื” ื™ื•ื›ืœ ืœืฉืžืฉ
10:41
to give virtual assistants a body and a face.
186
641434
3213
ื›ื“ื™ ืœืชืช ืœืขื•ื–ืจื™ื ื•ื™ืจื˜ื•ืืœื™ื™ื ื’ื•ืฃ ื•ืคื ื™ื.
10:44
A humanity.
187
644671
1192
ืื ื•ืฉื™ื•ืช.
10:45
I already love it that when I talk to virtual assistants,
188
645887
2762
ืื ื™ ื›ื‘ืจ ืื•ื”ื‘ ืืช ื–ื” ืฉื›ืฉืื ื™ ืžื“ื‘ืจ ืืœ ืขื•ื–ืจื™ื ื•ื™ืจื˜ื•ืืœื™ื™ื,
10:48
they answer back in a soothing, humanlike voice.
189
648673
2933
ื”ื ืขื•ื ื™ื ื‘ืงื•ืœ ืื ื•ืฉื™ ืžืจื’ื™ืข, .
10:51
Now they'll have a face.
190
651919
1776
ืขื›ืฉื™ื• ื™ื”ื™ื” ืœื”ื ืคืจืฆื•ืฃ.
10:53
And you'll get all the nonverbal cues that make communication so much easier.
191
653719
4882
ื•ืชืงื‘ืœื• ืืช ื›ืœ ื”ืจืžื–ื™ื ื”ืœื ืžื™ืœื•ืœื™ื™ื ืฉื”ื•ืคื›ื™ื ืชืงืฉื•ืจืช ืœื”ืจื‘ื” ื™ื•ืชืจ ื ื•ื—ื”.
11:00
It's going to be really nice.
192
660171
1420
ื–ื” ื™ื”ื™ื” ืžืžืฉ ื ื—ืžื“.
11:01
You'll be able to tell when a virtual assistant is busy or confused
193
661615
3637
ืชื•ื›ืœื• ืœื“ืขืช ืžืชื™ ืขื•ื–ืจ ื•ื™ืจื˜ื•ืืœื™ ืขืกื•ืง ืื• ืžื‘ื•ืœื‘ืœ
11:05
or concerned about something.
194
665276
2680
ืื• ืžื•ื“ืื’ ืœื’ื‘ื™ ืžืฉื”ื•.
11:09
Now, I couldn't leave the stage
195
669694
2509
ื•ืœื ื™ื›ื•ืœืชื™ ืœืขื–ื•ื‘ ืืช ื”ื‘ืžื”
11:12
without you actually being able to see my real face,
196
672227
2698
ืžื‘ืœื™ ืฉืชื•ื›ืœื• ืœืจืื•ืช ืืช ืคื ื™ ื”ืืžื™ืชื™ื™ื,
11:14
so you can do some comparison.
197
674949
1684
ื›ืš ืฉืืชื ื™ื›ื•ืœื™ื ืงืฆืช ืœื”ืฉื•ื•ืช,
11:18
So let me take off my helmet here.
198
678573
1876
ืื– ื”ืจืฉื• ืœื™ ืœื”ื•ืจื™ื“ ืืช ื”ืงืกื“ื” ืฉืœื™ ื›ืืŸ.
11:20
Yeah, don't worry, it looks way worse than it feels.
199
680473
4770
ืืœ ืชื—ืฉืฉื•, ื–ื” ื ืจืื” ื™ื•ืชืจ ื’ืจื•ืข ืžืžื” ืฉื–ื” ืžืจื’ื™ืฉ
11:25
(Laughter)
200
685267
2428
(ืฆื—ื•ืง)
11:29
So this is where we are.
201
689188
1698
ืื– ื–ื” ื”ืžืงื•ื ืฉื‘ื• ืื ื—ื ื• ื ืžืฆืื™ื.
11:30
Let me put this back on here.
202
690910
1604
ื”ืจืฉื• ืœื™ ืœื”ื—ื–ื™ืจ ืืช ื–ื” ืœื›ืืŸ.
11:32
(Laughter)
203
692538
1950
(ืฆื—ื•ืง)
11:35
Doink!
204
695115
1186
ืœืขื–ืื–ืœ!
11:37
So this is where we are.
205
697292
1600
ืื– ืคื” ืื ื—ื ื• ื ืžืฆืื™ื.
11:39
We're on the cusp of being able to interact with digital humans
206
699997
3701
ืื ื—ื ื• ืขืœ ืกืฃ ื”ื™ื›ื•ืœืช ืœืชืงืฉืจ ืขื ื‘ื ื™ ืื“ื ื“ื™ื’ื™ื˜ืœื™ื™ื
11:43
that are strikingly real,
207
703722
2181
ืฉื”ื ืืžื™ืชื™ื™ื ืœื”ืคืœื™ื,
11:45
whether they're being controlled by a person or a machine.
208
705927
3269
ื‘ื™ืŸ ืื ื”ื ื ืฉืœื˜ื™ื ืขืœ ื™ื“ื™ ืื“ื ืื• ืžื›ื•ื ื”.
11:49
And like all new technology these days,
209
709220
4375
ื•ื›ืžื• ื›ืœ ื˜ื›ื ื•ืœื•ื’ื™ื” ื—ื“ืฉื” ื‘ื™ืžื™ื ืืœื”,
11:54
it's going to come with some serious and real concerns
210
714531
4746
ื–ื” ื™ื‘ื•ื ืขื ื›ืžื” ื‘ืขื™ื•ืช ืจืฆื™ื ื™ื•ืช ื•ืžืžืฉื™ื•ืช
11:59
that we have to deal with.
211
719301
1734
ืฉืื ื• ืฆืจื™ื›ื™ื ืœื”ืชืžื•ื“ื“ ืื™ืชืŸ.
12:02
But I am just so really excited
212
722017
2118
ืื‘ืœ ืื ื™ ืคืฉื•ื˜ ื›ืœ ื›ืš ื ืœื”ื‘
12:04
about the ability to bring something that I've seen only in science fiction
213
724159
5053
ืžื”ื™ื›ื•ืœืช ืœื”ื‘ื™ื ืžืฉื”ื• ืฉืจืื™ืชื™ ืจืง ื‘ืžื“ืข ื‘ื“ื™ื•ื ื™
12:09
for my entire life
214
729236
2270
ื‘ืžืฉืš ื›ืœ ื—ื™ื™
12:11
into reality.
215
731530
1328
ืœืžืฆื™ืื•ืช.
12:13
Communicating with computers will be like talking to a friend.
216
733752
4222
ืชืงืฉื•ืจืช ืขื ืžื—ืฉื‘ื™ื ืชื”ื™ื” ื›ืžื• ืœื“ื‘ืจ ืขื ื—ื‘ืจ.
12:18
And talking to faraway friends
217
738323
2500
ื•ืœื“ื‘ืจ ืขื ื—ื‘ืจื™ื ื‘ืžืจื—ืงื™ื
12:20
will be like sitting with them together in the same room.
218
740847
3273
ื™ื”ื™ื” ื›ืžื• ืœืฉื‘ืช ืื™ืชื ื‘ืื•ืชื• ื—ื“ืจ.
12:24
Thank you very much.
219
744974
1308
ืชื•ื“ื” ืจื‘ื” ืœื›ื
12:26
(Applause)
220
746306
6713
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7