TED2018
Supasorn Suwajanakorn: Fake videos of real people -- and how to spot them
Supasorn Suwajanakorn: 真人的假视频以及如何识别他们
Filmed:
Readability: 4.6
1,453,308 views
你觉得你擅长发现假视频吗，当名人们并没有真正说过这些话的时候？这个演讲和技术演示让我们了解到这是如何做到的。计算机科学家 Supasorn Suwajanakorn 为我们展示了，如何使用人工智能和3D建模并与音频同步来创造假视频。了解更多关于这个技术的伦理问题和这个技术的可能性，以及我们应该采取的措施来对抗该技术的错误使用。
Supasorn Suwajanakorn - Computer scientist
Supasorn Suwajanakorn works on ways to reconstruct, preserve and reanimate anyone -- just from their existing photos and videos. Full bio
Supasorn Suwajanakorn works on ways to reconstruct, preserve and reanimate anyone -- just from their existing photos and videos. Full bio
Double-click the English transcript below to play the video.
00:12
Look at these images.
0
876
1151
看看这些图像。
00:14
Now, tell me which Obama here is real.
1
2051
2635
现在，告诉我哪个是真的奥马巴。
00:16
(Video) Barack Obama: To help families
refinance their homes,
refinance their homes,
2
4710
2861
巴拉克·奥巴马：帮助家庭对他们的房屋重做贷款，
00:19
to invest in things
like high-tech manufacturing,
like high-tech manufacturing,
3
7595
2647
投资高科技制造业，
00:22
clean energy
4
10266
1159
清洁能源
00:23
and the infrastructure
that creates good new jobs.
that creates good new jobs.
5
11449
2779
和带来良好就业机会的基础设施。
00:26
Supasorn Suwajanakorn: Anyone?
6
14647
1484
有人知道吗？
00:28
The answer is none of them.
7
16155
1874
答案是：都不是。
00:30
(Laughter)
8
18053
1114
（笑声）
00:31
None of these is actually real.
9
19191
1786
这些都不是真的。
00:33
So let me tell you how we got here.
10
21001
1840
那让我来告诉你们是怎么回事。
00:35
My inspiration for this work
11
23940
1578
我这个工作的灵感来自于
00:37
was a project meant to preserve our last
chance for learning about the Holocaust
chance for learning about the Holocaust
12
25542
5411
一个试图保存我们从幸存者那里
了解到的关于大屠杀
了解到的关于大屠杀
00:42
from the survivors.
13
30977
1768
的项目。
00:44
It's called New Dimensions in Testimony,
14
32769
2627
这个项目叫做证词新维度
(New Dimensions in Testimony)，
(New Dimensions in Testimony)，
00:47
and it allows you to have
interactive conversations
interactive conversations
15
35420
3126
它可以让你与真实大屠杀幸存者的全息图
00:50
with a hologram
of a real Holocaust survivor.
of a real Holocaust survivor.
16
38570
2556
进行互动对话。
00:53
(Video) Man: How did you
survive the Holocaust?
survive the Holocaust?
17
41793
1966
你是怎么在大屠杀中幸存下来的？
00:55
(Video) Hologram: How did I survive?
18
43783
1668
我怎么幸存下来？
00:57
I survived,
19
45912
1807
我幸存下来，
01:00
I believe,
20
48419
1527
我相信，
01:01
because providence watched over me.
21
49970
3023
是因为上帝眷顾我。
01:05
SS: Turns out these answers
were prerecorded in a studio.
were prerecorded in a studio.
22
53573
3454
原来这些答案是预先在工作室录制的。
01:09
Yet the effect is astounding.
23
57051
2452
但效果令人吃惊。
01:11
You feel so connected to his story
and to him as a person.
and to him as a person.
24
59527
3619
你会对他的故事，
他这个人感同身受。
他这个人感同身受。
01:16
I think there's something special
about human interaction
about human interaction
25
64011
3301
我想人类互动的特别之处
01:19
that makes it much more profound
26
67336
2757
让它比图书，演讲或电影
01:22
and personal
27
70117
2198
告诉我们的
01:24
than what books or lectures
or movies could ever teach us.
or movies could ever teach us.
28
72339
3485
要更加深刻和真实。
01:28
So I saw this and began to wonder,
29
76267
2425
所以我就开始想，
01:30
can we create a model
like this for anyone?
like this for anyone?
30
78716
2810
我们能不能为每个人做个模型？
01:33
A model that looks, talks
and acts just like them?
and acts just like them?
31
81550
2975
这个模型的样子，
谈话和举止就跟真人无异。
谈话和举止就跟真人无异。
01:37
So I set out to see if this could be done
32
85573
2007
于是我开始探索这个能不能搞定，
01:39
and eventually came up with a new solution
33
87604
2310
并最终找到了一个新的解决方案，
01:41
that can build a model of a person
using nothing but these:
using nothing but these:
34
89938
3220
只需使用下面这些东西就能构建人的模型：
01:45
existing photos and videos of a person.
35
93747
2214
个人现存的照片和视频。
01:48
If you can leverage
this kind of passive information,
this kind of passive information,
36
96701
2617
如果你能利用这种被动信息，
01:51
just photos and video that are out there,
37
99342
2007
只需公开的照片和视频，
01:53
that's the key to scaling to anyone.
38
101373
2056
这是扩展到其他人的关键。
01:56
By the way, here's Richard Feynman,
39
104119
1777
顺便说一句，这是理查德·费曼，
01:57
who in addition to being
a Nobel Prize winner in physics
a Nobel Prize winner in physics
40
105920
3413
他除了是诺贝尔物理学奖得主
02:01
was also known as a legendary teacher.
41
109357
2453
也是位传奇教师。
02:05
Wouldn't it be great
if we could bring him back
if we could bring him back
42
113080
2198
这岂不是很棒？
如果能够把他带回来
如果能够把他带回来
02:07
to give his lectures
and inspire millions of kids,
and inspire millions of kids,
43
115302
3265
讲课并激励成千上万的小孩，
02:10
perhaps not just in English
but in any language?
but in any language?
44
118591
2992
用英语或者其他任何语言？
02:14
Or if you could ask our grandparents
for advice and hear those comforting words
for advice and hear those comforting words
45
122441
4602
或者你也可以征求祖父母的意见，
听听那些让人宽慰的言语，
听听那些让人宽慰的言语，
02:19
even if they're no longer with us?
46
127067
1770
即便他们已经离开我们了。
02:21
Or maybe using this tool,
book authors, alive or not,
book authors, alive or not,
47
129683
3396
或者使用这个工具，图书的作者，
不管是活着的还是去世的，
不管是活着的还是去世的，
02:25
could read aloud all of their books
for anyone interested.
for anyone interested.
48
133103
2937
可以为任何有兴趣的人朗读他们的书本。
02:29
The creative possibilities
here are endless,
here are endless,
49
137199
2437
这里的创意可能是无限的，
02:31
and to me, that's very exciting.
50
139660
1713
对我而言，这非常让人兴奋。
02:34
And here's how it's working so far.
51
142595
2002
这是目前它的工作原理。
02:36
First, we introduce a new technique
52
144621
1667
首先我们引入一种新的技术
02:38
that can reconstruct a high-detailed
3D face model from any image
3D face model from any image
53
146312
4572
可以从任何图像中
重建一个高细节的3D人脸模型，
重建一个高细节的3D人脸模型，
02:42
without ever 3D-scanning the person.
54
150908
2119
而且无需经对真人进行3D扫描。
02:45
And here's the same output model
from different views.
from different views.
55
153890
2642
这是不同视角下的同一输出模型。
02:49
This also works on videos,
56
157969
1502
这也可以应用于视频，
02:51
by running the same algorithm
on each video frame
on each video frame
57
159495
2852
通过对每一幅视频
使用同样的算法
使用同样的算法
02:54
and generating a moving 3D model.
58
162371
2222
产生移动的3D模型。
02:57
And here's the same
output model from different angles.
output model from different angles.
59
165538
2772
这是不同视角下的同一输出模型。
03:01
It turns out this problem
is very challenging,
is very challenging,
60
169933
2534
这些问题富有挑战性，
但关键技巧在于我们需要提前
03:04
but the key trick
is that we are going to analyze
is that we are going to analyze
61
172491
2525
03:07
a large photo collection
of the person beforehand.
of the person beforehand.
62
175040
2966
分析一个人的大量照片集。
03:10
For George W. Bush,
we can just search on Google,
we can just search on Google,
63
178650
2539
对乔治·沃克·布什，
我们只需要搜索谷歌，
我们只需要搜索谷歌，
03:14
and from that, we are able
to build an average model,
to build an average model,
64
182309
2499
这样，我们就能建立一个平均模型，
03:16
an iterative, refined model
to recover the expression
to recover the expression
65
184832
3111
一个迭代，精炼的模型来恢复表达的细节，
03:19
in fine details,
like creases and wrinkles.
like creases and wrinkles.
66
187967
2336
比如折痕和皱纹。
03:23
What's fascinating about this
67
191326
1403
迷人的是
03:24
is that the photo collection
can come from your typical photos.
can come from your typical photos.
68
192753
3423
照片集可以来自你的特定照片。
03:28
It doesn't really matter
what expression you're making
what expression you're making
69
196200
2603
你做何表情或者你在哪里拍照
03:30
or where you took those photos.
70
198827
1885
并不那么关键。
03:32
What matters is
that there are a lot of them.
that there are a lot of them.
71
200736
2400
关键的是数量要足够多。
03:35
And we are still missing color here,
72
203160
1736
这里我们仍然缺少肤色，
03:36
so next, we develop
a new blending technique
a new blending technique
73
204920
2348
所以下一步，
我们开发了一种新的混合技术
我们开发了一种新的混合技术
03:39
that improves upon
a single averaging method
a single averaging method
74
207292
2836
改善了平均模型，
03:42
and produces sharp
facial textures and colors.
facial textures and colors.
75
210152
2818
并产生尖锐的面部纹理和肤色。
03:45
And this can be done for any expression.
76
213779
2771
这可以用于做任何表情。
03:49
Now we have a control
of a model of a person,
of a model of a person,
77
217485
2499
现在我们可以
对一个人的模型进行控制，
对一个人的模型进行控制，
03:52
and the way it's controlled now
is by a sequence of static photos.
is by a sequence of static photos.
78
220008
3795
它现在被控制的方式是
一系列静态的照片。
一系列静态的照片。
03:55
Notice how the wrinkles come and go,
depending on the expression.
depending on the expression.
79
223827
3126
注意皱纹是如何产生和消失的，
这取决于你的表情。
这取决于你的表情。
04:00
We can also use a video
to drive the model.
to drive the model.
80
228109
2746
我们也可以使用视频来驱动模型。
04:02
(Video) Daniel Craig: Right, but somehow,
81
230879
2593
丹尼尔·克雷格：没错，但不管怎样，
04:05
we've managed to attract
some more amazing people.
some more amazing people.
82
233496
3771
我们能够吸引到更多优秀的人才。
04:10
SS: And here's another fun demo.
83
238021
1642
这是另一个有趣的演示。
04:11
So what you see here
are controllable models
are controllable models
84
239687
2246
所以你们看到的是
我使用人们的互联网图像
我使用人们的互联网图像
04:13
of people I built
from their internet photos.
from their internet photos.
85
241957
2444
建立的个人控制模型。
04:16
Now, if you transfer
the motion from the input video,
the motion from the input video,
86
244425
2904
现在，如果你从视频中传递表情动作，
04:19
we can actually drive the entire party.
87
247353
2152
我们可以让整个派对动起来。
04:21
George W. Bush:
It's a difficult bill to pass,
It's a difficult bill to pass,
88
249529
2172
布什：这是个难以通过的法案，
04:23
because there's a lot of moving parts,
89
251725
2303
因为有太多可供商榷的部分，
04:26
and the legislative processes can be ugly.
90
254052
5231
立法过程可能让人奔溃。
04:31
(Applause)
91
259307
1630
（鼓掌）
04:32
SS: So coming back a little bit,
92
260961
1837
那么回到正题，
04:34
our ultimate goal, rather,
is to capture their mannerisms
is to capture their mannerisms
93
262822
3191
我们的最终目标，
不如说，是捕捉他们的言谈举止，
不如说，是捕捉他们的言谈举止，
04:38
or the unique way each
of these people talks and smiles.
of these people talks and smiles.
94
266037
3045
或者每一个人交谈或微笑的独特之处。
04:41
So to do that, can we
actually teach the computer
actually teach the computer
95
269106
2313
所以这样，
我们能不能只向电脑展示这个人的录像
我们能不能只向电脑展示这个人的录像
04:43
to imitate the way someone talks
96
271443
2222
就能教会电脑
04:45
by only showing it
video footage of the person?
video footage of the person?
97
273689
2420
去模仿人们谈话的方式？
04:48
And what I did exactly was,
I let a computer watch
I let a computer watch
98
276898
2577
而我做的事情是，我让电脑
04:51
14 hours of pure Barack Obama
giving addresses.
giving addresses.
99
279499
3277
看了14个小时的奥巴马演讲。
04:55
And here's what we can produce
given only his audio.
given only his audio.
100
283443
3516
这是我们只通过他的音频生产出来的内容。
04:58
(Video) BO: The results are clear.
101
286983
1777
结果非常明显。
05:00
America's businesses have created
14.5 million new jobs
14.5 million new jobs
102
288784
4349
在过去75个月中，美国企业已经创造了
05:05
over 75 straight months.
103
293157
2774
1450万新的工作机会。
05:07
SS: So what's being synthesized here
is only the mouth region,
is only the mouth region,
104
295955
2905
所以这里合成的只是嘴巴部分，
05:10
and here's how we do it.
105
298884
1540
这是我们做的方法。
05:12
Our pipeline uses a neural network
106
300764
1826
我们的处理系统使用神经网络
05:14
to convert and input audio
into these mouth points.
into these mouth points.
107
302614
2936
来转换和输入音频到这些嘴巴的位置。
05:18
(Video) BO: We get it through our job
or through Medicare or Medicaid.
or through Medicare or Medicaid.
108
306547
4225
我们通过我们的工作或者医疗保险
或补助来实现这一目标。
或补助来实现这一目标。
05:22
SS: Then we synthesize the texture,
enhance details and teeth,
enhance details and teeth,
109
310796
3420
然后我们合成纹理，
增强细节和牙齿，
增强细节和牙齿，
05:26
and blend it into the head
and background from a source video.
and background from a source video.
110
314240
3074
并将其与源视频中的
头部和背景混合在一起。
头部和背景混合在一起。
05:29
(Video) BO: Women can get free checkups,
111
317338
1905
女性可以获得免费的检查，
05:31
and you can't get charged more
just for being a woman.
just for being a woman.
112
319267
2968
你不会因为是女性而需要支付更高的费用。
05:34
Young people can stay
on a parent's plan until they turn 26.
on a parent's plan until they turn 26.
113
322973
3306
年轻人可以在父母计划中呆到26岁。
05:39
SS: I think these results
seem very realistic and intriguing,
seem very realistic and intriguing,
114
327267
2952
我觉得这些结果看起来非常真实和有趣，
05:42
but at the same time
frightening, even to me.
frightening, even to me.
115
330243
3173
但同时，也让我担忧，即便是我。
05:45
Our goal was to build an accurate model
of a person, not to misrepresent them.
of a person, not to misrepresent them.
116
333440
4015
我们的目标是构建人的精准模型，
而非歪曲他们。
而非歪曲他们。
05:49
But one thing that concerns me
is its potential for misuse.
is its potential for misuse.
117
337956
3111
但让我担忧的是它被错误使用的可能。
05:53
People have been thinking
about this problem for a long time,
about this problem for a long time,
118
341958
2971
人们思考这个问题很长时间了，
05:56
since the days when Photoshop
first hit the market.
first hit the market.
119
344953
2381
从Photoshop进入市场那天就开始了。
05:59
As a researcher, I'm also working
on countermeasure technology,
on countermeasure technology,
120
347862
3801
作为一名研究人员，
我也在研究对抗技术，
我也在研究对抗技术，
06:03
and I'm part of an ongoing
effort at AI Foundation,
effort at AI Foundation,
121
351687
2942
我是人工智能基金会持续努力的一份子，
06:06
which uses a combination
of machine learning and human moderators
of machine learning and human moderators
122
354653
3397
它结合了机器学习和人工模型
06:10
to detect fake images and videos,
123
358074
2144
来识别假图像和视频，
06:12
fighting against my own work.
124
360242
1514
与我们自己的工作做斗争。
06:14
And one of the tools we plan to release
is called Reality Defender,
is called Reality Defender,
125
362675
3190
我们打算发布的一个工具叫做真相卫士，
06:17
which is a web-browser plug-in
that can flag potentially fake content
that can flag potentially fake content
126
365889
4039
是个浏览器插件
可以用来自动标记潜在假内容，
可以用来自动标记潜在假内容，
06:21
automatically, right in the browser.
127
369952
2533
在浏览器中就可以使用。
06:24
(Applause)
128
372509
4228
（掌声）
06:28
Despite all this, though,
129
376761
1453
此外，
06:30
fake videos could do a lot of damage,
130
378238
1840
假视频可以带来很大危害，
06:32
even before anyone has a chance to verify,
131
380102
3294
甚至在人们有机会验证它之前，
06:35
so it's very important
that we make everyone aware
that we make everyone aware
132
383420
2722
所以让大家意识到这可能是什么
06:38
of what's currently possible
133
386166
2007
非常重要，
06:40
so we can have the right assumption
and be critical about what we see.
and be critical about what we see.
134
388197
3369
这样我们才能得到正确的推断，
并对看到的保持谨慎。
并对看到的保持谨慎。
06:44
There's still a long way to go before
we can fully model individual people
we can fully model individual people
135
392423
5007
在个人完全建模
以及确保技术的安全性方面，
以及确保技术的安全性方面，
06:49
and before we can ensure
the safety of this technology.
the safety of this technology.
136
397454
2786
仍有很长的路要走。
06:53
But I'm excited and hopeful,
137
401097
1587
但我兴奋且充满希望，
06:54
because if we use it right and carefully,
138
402708
3539
因为如果我们正确地使用它，
06:58
this tool can allow any individual's
positive impact on the world
positive impact on the world
139
406271
4309
这个工具可以让
每个人对世界积极的影响
每个人对世界积极的影响
07:02
to be massively scaled
140
410604
2190
得到大规模的普及
07:04
and really help shape our future
the way we want it to be.
the way we want it to be.
141
412818
2742
并真正帮助塑造我们想要的未来。
07:07
Thank you.
142
415584
1151
谢谢。
07:08
(Applause)
143
416759
5090
（掌声）
ABOUT THE SPEAKER
Supasorn Suwajanakorn - Computer scientistSupasorn Suwajanakorn works on ways to reconstruct, preserve and reanimate anyone -- just from their existing photos and videos.
Why you should listen
Can we create a digital avatar that looks, acts and talks just like our sweet grandma? This question has inspired Supasorn Suwajanakorn, a recent PhD graduate from the University of Washington, to spend years developing new tools to make it a reality. He has developed a set of algorithms that can build a moving 3D face model of anyone from just photos, which was awarded the Innovation of the Year in 2016. He then introduced the first system that can replicate a person's speech and produce a realistic CG-animation by only analyzing their existing video footage -- all without ever bringing in the person to a Hollywood capture studio.
Suwajanakorn is working in the field of machine learning and computer vision. His goal is to bring vision algorithms out of the lab and make them work in the wild.
Supasorn Suwajanakorn | Speaker | TED.com