1
00:00:00,556 --> 00:00:04,573
Our emotions influence
every aspect of our lives,

2
00:00:04,573 --> 00:00:08,149
from our health and how we learn
to how we do business and make decisions,

3
00:00:08,149 --> 00:00:10,262
big ones and small.

4
00:00:10,262 --> 00:00:14,162
Our emotions also influence
how we connect with one another.

5
00:00:14,162 --> 00:00:19,108
We've evolved to live
in a world like this,

6
00:00:19,108 --> 00:00:23,427
but instead, we're living
more and more of our lives like this.

7
00:00:23,427 --> 00:00:26,561
-- This is the text message
from my daughter last night --

8
00:00:26,561 --> 00:00:29,301
in a world that's devoid of emotion.

9
00:00:29,301 --> 00:00:31,252
So I'm on a mission to change that.

10
00:00:31,252 --> 00:00:35,733
I want to bring emotions
back into our digital experiences.

11
00:00:35,733 --> 00:00:39,030
I started on this path 15 years ago.

12
00:00:39,030 --> 00:00:41,366
I was a computer scientist in Egypt,

13
00:00:41,366 --> 00:00:45,871
and I had just gotten accepted
to a Ph.D program at Cambridge University.

14
00:00:45,871 --> 00:00:47,984
So, I did something quite unusual

15
00:00:47,984 --> 00:00:52,209
for a young newlywed Muslim Egyptian wife.

16
00:00:52,209 --> 00:00:56,598
With the support of my husband,
who had to stay in Egypt,

17
00:00:56,598 --> 00:00:59,616
I packed my bags and I moved to England.

18
00:00:59,616 --> 00:01:02,844
At Cambridge, thousands of miles
away from home,

19
00:01:02,844 --> 00:01:06,257
I realized I was spending
more hours with my laptop

20
00:01:06,257 --> 00:01:08,486
than I did with any other human.

21
00:01:08,486 --> 00:01:13,339
Yet despite this intimacy, my laptop
had absolutely no idea how I was feeling.

22
00:01:13,339 --> 00:01:16,055
It had no idea if I was happy,

23
00:01:16,055 --> 00:01:19,538
having a bad day, or stressed, confused,

24
00:01:19,538 --> 00:01:22,046
and so that got frustrating.

25
00:01:22,046 --> 00:01:28,501
Even worse, as I communicated
online with my family back home,

26
00:01:28,501 --> 00:01:32,703
I felt that all my emotions
disappeared in cyberspace.

27
00:01:32,703 --> 00:01:37,858
I was homesick, I was lonely,
and on some days I was actually crying,

28
00:01:37,858 --> 00:01:42,786
but all I had to communicate
these emotions was this.

29
00:01:42,786 --> 00:01:44,806
(Laughter)

30
00:01:44,806 --> 00:01:49,078
Today's technology
has lots of IQ, but no EQ;

31
00:01:49,078 --> 00:01:52,956
lots of cognitive intelligence,
but no emotional intelligence.

32
00:01:52,956 --> 00:01:55,153
So that got me thinking, you know,

33
00:01:55,153 --> 00:01:58,777
what if our technology
could sense our emotions?

34
00:01:58,777 --> 00:02:02,853
What if our devices could sense
how we felt and reacted accordingly,

35
00:02:02,853 --> 00:02:06,666
just the way an emotionally
intelligent friend would?

36
00:02:06,666 --> 00:02:10,079
Those questions led me and my team

37
00:02:10,079 --> 00:02:14,607
to create technologies that can read
and respond to our emotions,

38
00:02:14,607 --> 00:02:18,577
and our starting point was the human face.

39
00:02:18,577 --> 00:02:21,750
So our human face happens to be
one of the most powerful channels

40
00:02:21,750 --> 00:02:25,766
that we all use to communicate
social and emotional state,

41
00:02:25,766 --> 00:02:28,776
everything from enjoyment, surprise,

42
00:02:28,776 --> 00:02:32,979
empathy, and curiosity.

43
00:02:32,979 --> 00:02:37,907
In emotion science, we call each
facial muscle movement an action unit.

44
00:02:37,907 --> 00:02:40,832
So for example, action unit 12,

45
00:02:40,832 --> 00:02:43,015
it's not a Hollywood blockbuster,

46
00:02:43,015 --> 00:02:46,312
it is actually a lip corner pull,
which is the main component of a smile.

47
00:02:46,312 --> 00:02:49,052
Try it everybody. Let's get
some smiles going on.

48
00:02:49,052 --> 00:02:51,954
Another example is action unit 4.
It's the brow furrow.

49
00:02:51,954 --> 00:02:54,392
It's when you draw your eyebrows together

50
00:02:54,392 --> 00:02:56,459
and you create all
these textures and wrinkles.

51
00:02:56,459 --> 00:03:00,754
We don't like them, but it's
a strong indicator of a negative emotion.

52
00:03:00,754 --> 00:03:02,960
So we have about 45 of these action units,

53
00:03:02,960 --> 00:03:06,350
and they combine to express
hundreds of emotions.

54
00:03:06,350 --> 00:03:10,251
Teaching a computer to read
these facial emotions is hard,

55
00:03:10,251 --> 00:03:13,223
because these action units,
they can be fast, they're subtle,

56
00:03:13,223 --> 00:03:15,777
and they combine in many different ways.

57
00:03:15,777 --> 00:03:19,515
So take, for example,
the smile and the smirk.

58
00:03:19,515 --> 00:03:23,268
They look somewhat similar,
but they mean very different things.

59
00:03:23,268 --> 00:03:24,986
(Laughter)

60
00:03:24,986 --> 00:03:27,099
So the smile is positive,

61
00:03:27,099 --> 00:03:29,026
a smirk is often negative.

62
00:03:29,026 --> 00:03:33,136
Sometimes a smirk
can make you became famous.

63
00:03:33,136 --> 00:03:35,960
But seriously, it's important
for a computer to be able

64
00:03:35,960 --> 00:03:38,815
to tell the difference
between the two expressions.

65
00:03:38,815 --> 00:03:40,627
So how do we do that?

66
00:03:40,627 --> 00:03:42,414
We give our algorithms

67
00:03:42,414 --> 00:03:46,524
tens of thousands of examples
of people we know to be smiling,

68
00:03:46,524 --> 00:03:49,589
from different ethnicities, ages, genders,

69
00:03:49,589 --> 00:03:52,004
and we do the same for smirks.

70
00:03:52,004 --> 00:03:53,954
And then, using deep learning,

71
00:03:53,954 --> 00:03:56,810
the algorithm looks for all these
textures and wrinkles

72
00:03:56,810 --> 00:03:59,039
and shape changes on our face,

73
00:03:59,039 --> 00:04:02,592
and basically learns that all smiles
have common characteristics,

74
00:04:02,592 --> 00:04:05,773
all smirks have subtly
different characteristics.

75
00:04:05,773 --> 00:04:08,141
And the next time it sees a new face,

76
00:04:08,141 --> 00:04:10,440
it essentially learns that, you know,

77
00:04:10,440 --> 00:04:13,473
this face has the same
characteristics of a smile,

78
00:04:13,473 --> 00:04:17,791
and it says, "Aha, I recognize this.
This is a smile expression."

79
00:04:17,791 --> 00:04:21,181
So the best way to demonstrate
how this technology works

80
00:04:21,181 --> 00:04:23,317
is to try a live demo,

81
00:04:23,317 --> 00:04:24,850
so I need a volunteer,

82
00:04:24,850 --> 00:04:27,195
preferably somebody with a face.

83
00:04:27,195 --> 00:04:29,564
(Laughter)

84
00:04:29,564 --> 00:04:32,335
Chloe's going to be our volunteer today.

85
00:04:33,325 --> 00:04:37,783
So over the past five years, we've moved
from being a research project at MIT

86
00:04:37,783 --> 00:04:39,269
to a company,

87
00:04:39,269 --> 00:04:41,971
where my team has worked really hard
to make this technology work,

88
00:04:41,971 --> 00:04:44,075
as we like to say, in the wild.

89
00:04:44,075 --> 00:04:47,210
And we've also shrunk it so that
the core emotion engine

90
00:04:47,210 --> 00:04:50,530
works on any mobile device
with a camera, like this iPad.

91
00:04:50,530 --> 00:04:54,756
So let's give this a try.

92
00:04:54,756 --> 00:04:58,680
As you can see, the algorithm
has essentially found Chloe's face,

93
00:04:58,680 --> 00:05:00,552
so it's this white bounding box,

94
00:05:00,552 --> 00:05:02,943
and it's tracking the main
feature points on her face,

95
00:05:02,943 --> 00:05:05,799
so her eyebrows, her eyes,
her mouth, and her nose.

96
00:05:05,799 --> 00:05:08,786
The question is,
can it recognize her expression?

97
00:05:08,786 --> 00:05:10,457
So we're going to test the machine.

98
00:05:10,457 --> 00:05:14,823
So first of all, give me your poker face.
Yep, awesome. (Laughter)

99
00:05:14,823 --> 00:05:17,586
And then as she smiles,
this is a genuine smile, it's great.

100
00:05:17,586 --> 00:05:19,968
So you can see the green bar
go up as she smiles.

101
00:05:19,968 --> 00:05:22,151
Now that was a big smile.
Can you try, like, a subtle smile

102
00:05:22,151 --> 00:05:25,402
to see if the computer can recognize?
It does recognize subtle smiles as well.

103
00:05:25,402 --> 00:05:27,607
We've worked really hard
to make that happen.

104
00:05:27,607 --> 00:05:31,439
And then eyebrow raised,
indicator of surprise.

105
00:05:31,439 --> 00:05:35,688
Brow furrow, which is
an indicator of confusion.

106
00:05:35,688 --> 00:05:39,565
Frown. Yes, perfect.

107
00:05:39,565 --> 00:05:42,468
So these are all the different
action units. There's many more of them.

108
00:05:42,468 --> 00:05:45,022
This is just a slimmed down demo.

109
00:05:45,022 --> 00:05:48,528
But we call each reading
an emotion data point,

110
00:05:48,528 --> 00:05:51,337
and then they can fire together
to portray different emotions.

111
00:05:51,337 --> 00:05:55,099
So on the right side of the demo,
like, look like you're happy.

112
00:05:55,099 --> 00:05:57,444
So that's joy. Joy fires up.

113
00:05:57,444 --> 00:05:59,371
And then give me a disgust face.

114
00:05:59,371 --> 00:06:03,643
Try to remember what it was like
when Zayn left One Direction.

115
00:06:03,643 --> 00:06:05,153
(Laughter)

116
00:06:05,153 --> 00:06:09,495
Yeah, wrinkle your nose. Awesome.

117
00:06:09,495 --> 00:06:13,326
And the valance is actually quite
negative, so you must have been a big fan.

118
00:06:13,326 --> 00:06:15,926
So valance is how positive
or negative an experience is,

119
00:06:15,926 --> 00:06:18,712
and engagement is how
expressive she is as well.

120
00:06:18,712 --> 00:06:22,126
So imagine if Chloe had access
to this realtime emotion stream,

121
00:06:22,126 --> 00:06:24,935
and she could share it
with anybody she wanted to.

122
00:06:24,935 --> 00:06:27,858
Thank you.

123
00:06:27,858 --> 00:06:33,749
(Applause)

124
00:06:33,749 --> 00:06:35,979
So so far, we have amassed

125
00:06:35,979 --> 00:06:38,788
12 billion of these emotion data points.

126
00:06:38,788 --> 00:06:41,063
It's the largest emotion
database in the world.

127
00:06:41,063 --> 00:06:44,593
We've collected it
from 2.9 million face videos,

128
00:06:44,593 --> 00:06:47,193
people who have agreed
to share their emotions with us,

129
00:06:47,193 --> 00:06:50,398
and from 75 countries around the world.

130
00:06:50,398 --> 00:06:52,603
It's growing every day.

131
00:06:52,603 --> 00:06:54,670
It blows my mind away

132
00:06:54,670 --> 00:06:57,865
that we can now quantify something
as personal as our emotions,

133
00:06:57,865 --> 00:07:00,001
and we can do it at this scale.

134
00:07:00,001 --> 00:07:02,277
So what have we learned to date?

135
00:07:02,277 --> 00:07:05,388
Gender.

136
00:07:05,388 --> 00:07:09,034
Our data confirms something
that you might suspect.

137
00:07:09,034 --> 00:07:10,891
Women are more expressive than men.

138
00:07:10,891 --> 00:07:13,724
Not only do they smile more,
their smiles last longer,

139
00:07:13,724 --> 00:07:16,478
and we can now really quantify
what it is that men and women

140
00:07:16,478 --> 00:07:18,614
respond to differently.

141
00:07:18,614 --> 00:07:20,904
Let's do culture: so in the United States,

142
00:07:20,904 --> 00:07:24,108
women are 40 percent
more expressive than men,

143
00:07:24,108 --> 00:07:27,753
but curiously, we don't see any difference
in the U.K. between men and women.

144
00:07:27,753 --> 00:07:30,259
(Laughter)

145
00:07:31,445 --> 00:07:33,256
Age:

146
00:07:33,256 --> 00:07:35,323
people who are 50 years and older

147
00:07:35,323 --> 00:07:38,759
are 25 percent more emotive
than younger people.

148
00:07:38,759 --> 00:07:43,751
Women in their 20s smile a lot more
than men the same age,

149
00:07:43,751 --> 00:07:46,770
perhaps a necessity for dating,

150
00:07:46,770 --> 00:07:49,997
but perhaps what surprised us
the most about this data

151
00:07:49,997 --> 00:07:53,410
is that we happen
to be expressive all the time,

152
00:07:53,410 --> 00:07:56,243
even when we are sitting
in front of our devices alone,

153
00:07:56,243 --> 00:07:59,517
and it's not just when we're watching
cat videos on Facebook.

154
00:07:59,517 --> 00:08:02,857
We are expressive when we're emailing,
texting, shopping online,

155
00:08:02,857 --> 00:08:05,527
or even doing our taxes.

156
00:08:05,527 --> 00:08:07,919
Where is this data used today?

157
00:08:07,919 --> 00:08:10,682
In understanding how we engage with media,

158
00:08:10,682 --> 00:08:13,166
so understanding virality
and voting behavior;

159
00:08:13,166 --> 00:08:15,906
and also empowering
or emotion-enabling technology,

160
00:08:15,906 --> 00:08:20,527
and I want to share some examples
that are especially close to my heart.

161
00:08:20,527 --> 00:08:24,265
Emotion-enabled wearable glasses
can help individuals

162
00:08:24,265 --> 00:08:27,493
who are visually impaired
read the faces of others,

163
00:08:27,493 --> 00:08:31,068
and it can help individuals
on the autism spectrum interpret emotion,

164
00:08:31,068 --> 00:08:34,458
something that they really struggle with.

165
00:08:34,458 --> 00:08:38,777
In education, imagine if
your learning apps

166
00:08:38,777 --> 00:08:41,587
sense that you're confused and slow down,

167
00:08:41,587 --> 00:08:43,444
or that you're bored, so it sped up,

168
00:08:43,444 --> 00:08:46,973
just like a great teacher would
in a classroom.

169
00:08:46,973 --> 00:08:49,644
What if your wristwatch tracks your mood,

170
00:08:49,644 --> 00:08:52,337
or your car sensed that you're tired,

171
00:08:52,337 --> 00:08:54,885
or perhaps your fridge
knows that you're stressed,

172
00:08:54,885 --> 00:09:00,951
so it auto-locks to prevent you
from binge eating. (Laughter)

173
00:09:00,951 --> 00:09:03,668
I would like that, yeah.

174
00:09:03,668 --> 00:09:05,595
What if, when I was in Cambridge,

175
00:09:05,595 --> 00:09:07,908
I had access to my realtime
emotion stream,

176
00:09:07,908 --> 00:09:11,437
and I could share that with my family
back home in a very natural way,

177
00:09:11,437 --> 00:09:15,408
just like I would if we were all
in the same room together?

178
00:09:15,408 --> 00:09:18,055
I think in five years down the line,

179
00:09:18,055 --> 00:09:20,887
all our devices are going
to have an emotion chip,

180
00:09:20,887 --> 00:09:24,951
and we won't remember what it was like
when we couldn't just frown at our device

181
00:09:24,951 --> 00:09:29,200
and our device would say, "Hmm,
you didn't like that, did you."

182
00:09:29,200 --> 00:09:32,961
Our biggest challenge is that there are
so many applications of this technology,

183
00:09:32,961 --> 00:09:35,864
my team and I realize that we can't
build them all ourselves,

184
00:09:35,864 --> 00:09:39,036
so we've made this technology available
so that other developers

185
00:09:39,036 --> 00:09:41,474
can get building and get creative.

186
00:09:41,474 --> 00:09:45,560
We recognize that
there are potential risks

187
00:09:45,560 --> 00:09:47,627
and potential for abuse,

188
00:09:47,627 --> 00:09:50,576
but personally, having spent
many years doing this,

189
00:09:50,576 --> 00:09:53,548
I believe that the benefits to humanity

190
00:09:53,548 --> 00:09:55,823
from having emotionally
intelligent technology

191
00:09:55,823 --> 00:09:59,399
far outweigh the potential for misuse.

192
00:09:59,399 --> 00:10:01,930
And I invite you all to be
part of the conversation.

193
00:10:01,930 --> 00:10:04,484
The more people who know
about this technology,

194
00:10:04,484 --> 00:10:08,431
the more we can all have a voice
in how it's being used.

195
00:10:08,431 --> 00:10:13,655
So as more and more
of our lives become digital,

196
00:10:13,655 --> 00:10:17,153
we are fighting a losing battle
trying to curb our usage of devices

197
00:10:17,153 --> 00:10:19,382
in order to reclaim our emotions.

198
00:10:19,382 --> 00:10:24,536
So what I'm trying to do instead
is to bring emotions into our technology

199
00:10:24,536 --> 00:10:26,765
and make our technologies more responsive.

200
00:10:26,765 --> 00:10:29,435
So I want those devices
that have separated us

201
00:10:29,435 --> 00:10:31,897
to bring us back together,

202
00:10:31,897 --> 00:10:36,485
and by humanizing technology,
we have this golden opportunity

203
00:10:36,485 --> 00:10:39,782
to re-imagine how we
connect with machines,

204
00:10:39,782 --> 00:10:44,263
and therefore, how we, as human beings,

205
00:10:44,263 --> 00:10:46,167
connect with one another.

206
00:10:46,167 --> 00:10:48,327
Thank you.

207
00:10:48,327 --> 00:10:51,064
(Applause)