TED2015
Fei-Fei Li: How we're teaching computers to understand pictures
李飞飞: 我们怎么教计算机理解图片？
Filmed:
Readability: 4.5
2,702,344 views
当一个非常小的孩子看到图片时，她可以辨认出里面简单的要素："猫""书""椅子"。现在，电脑也聪明得可以做同样的工作了。接下来呢？在这个令人震撼的演讲里，计算机视觉专家李飞飞介绍了这一技术的发展现状——包括她的团队为了"教"计算机理解图片所建立的一千五百万照片的数据库——而关键性的要点还不止于此。
Fei-Fei Li - Computer scientist
As Director of Stanford’s Artificial Intelligence Lab and Vision Lab, Fei-Fei Li is working to solve AI’s trickiest problems -- including image recognition, learning and language processing. Full bio
As Director of Stanford’s Artificial Intelligence Lab and Vision Lab, Fei-Fei Li is working to solve AI’s trickiest problems -- including image recognition, learning and language processing. Full bio
Double-click the English transcript below to play the video.
00:14
Let me show you something.
0
2366
3738
我先来给你们看点东西。
（视频）女孩：
好吧，这是只猫，坐在床上。
好吧，这是只猫，坐在床上。
00:18
(Video) Girl: Okay, that's a cat
sitting in a bed.
sitting in a bed.
1
6104
4156
一个男孩摸着一头大象。
00:22
The boy is petting the elephant.
2
10260
4040
那些人正准备登机。
00:26
Those are people
that are going on an airplane.
that are going on an airplane.
3
14300
4354
00:30
That's a big airplane.
4
18654
2810
那是架大飞机。
李飞飞：
这是一个三岁的小孩
这是一个三岁的小孩
00:33
Fei-Fei Li: This is
a three-year-old child
a three-year-old child
5
21464
2206
在讲述她从一系列照片里看到的东西。
00:35
describing what she sees
in a series of photos.
in a series of photos.
6
23670
3679
00:39
She might still have a lot
to learn about this world,
to learn about this world,
7
27349
2845
对这个世界，
她也许还有很多要学的东西，
她也许还有很多要学的东西，
00:42
but she's already an expert
at one very important task:
at one very important task:
8
30194
4549
但在一个重要的任务上，
她已经是专家了：
她已经是专家了：
00:46
to make sense of what she sees.
9
34743
2846
去理解她所看到的东西。
我们的社会已经在科技上
取得了前所未有的进步。
取得了前所未有的进步。
00:50
Our society is more
technologically advanced than ever.
technologically advanced than ever.
10
38229
4226
00:54
We send people to the moon,
we make phones that talk to us
we make phones that talk to us
11
42455
3629
我们把人送上月球，
我们制造出可以与我们对话的手机，
我们制造出可以与我们对话的手机，
00:58
or customize radio stations
that can play only music we like.
that can play only music we like.
12
46084
4946
或者订制一个音乐电台，
播放的全是我们喜欢的音乐。
播放的全是我们喜欢的音乐。
01:03
Yet, our most advanced
machines and computers
machines and computers
13
51030
4055
然而，哪怕是我们最先进的机器和电脑
01:07
still struggle at this task.
14
55085
2903
也会在这个问题上犯难。
01:09
So I'm here today
to give you a progress report
to give you a progress report
15
57988
3459
所以今天我在这里，
向大家做个进度汇报：
向大家做个进度汇报：
01:13
on the latest advances
in our research in computer vision,
in our research in computer vision,
16
61447
4047
关于我们在计算机
视觉方面最新的研究进展。
视觉方面最新的研究进展。
01:17
one of the most frontier
and potentially revolutionary
and potentially revolutionary
17
65494
4161
这是计算机科学领域最前沿的、
01:21
technologies in computer science.
18
69655
3206
具有革命性潜力的科技。
01:24
Yes, we have prototyped cars
that can drive by themselves,
that can drive by themselves,
19
72861
4551
是的，我们现在已经有了
具备自动驾驶功能的原型车，
具备自动驾驶功能的原型车，
01:29
but without smart vision,
they cannot really tell the difference
they cannot really tell the difference
20
77412
3853
但是如果没有敏锐的视觉，
它们就不能真正区分出
它们就不能真正区分出
01:33
between a crumpled paper bag
on the road, which can be run over,
on the road, which can be run over,
21
81265
3970
地上摆着的是一个压扁的纸袋，
可以被轻易压过，
可以被轻易压过，
01:37
and a rock that size,
which should be avoided.
which should be avoided.
22
85235
3340
还是一块相同体积的石头，
应该避开。
应该避开。
01:41
We have made fabulous megapixel cameras,
23
89415
3390
我们已经造出了超高清的相机，
01:44
but we have not delivered
sight to the blind.
sight to the blind.
24
92805
3135
但我们仍然无法把
这些画面传递给盲人。
这些画面传递给盲人。
01:48
Drones can fly over massive land,
25
96420
3305
我们的无人机可以飞跃广阔的土地，
01:51
but don't have enough vision technology
26
99725
2134
却没有足够的视觉技术
去帮我们追踪热带雨林的变化。
01:53
to help us to track
the changes of the rainforests.
the changes of the rainforests.
27
101859
3461
01:57
Security cameras are everywhere,
28
105320
2950
安全摄像头到处都是，
02:00
but they do not alert us when a child
is drowning in a swimming pool.
is drowning in a swimming pool.
29
108270
5067
但当有孩子在泳池里溺水时
它们无法向我们报警。
它们无法向我们报警。
02:06
Photos and videos are becoming
an integral part of global life.
an integral part of global life.
30
114167
5595
照片和视频，已经成为
全人类生活里不可缺少的部分。
全人类生活里不可缺少的部分。
02:11
They're being generated at a pace
that's far beyond what any human,
that's far beyond what any human,
31
119762
4087
它们以极快的速度被创造出来，
以至于没有任何人，或者团体，
以至于没有任何人，或者团体，
02:15
or teams of humans, could hope to view,
32
123849
2783
能够完全浏览这些内容，
02:18
and you and I are contributing
to that at this TED.
to that at this TED.
33
126632
3921
而你我正参与其中的这场TED，
也为之添砖加瓦。
也为之添砖加瓦。
02:22
Yet our most advanced software
is still struggling at understanding
is still struggling at understanding
34
130553
5232
直到现在，我们最先进的
软件也依然为之犯难：
软件也依然为之犯难：
02:27
and managing this enormous content.
35
135785
3876
该怎么理解和处理
这些数量庞大的内容？
这些数量庞大的内容？
02:31
So in other words,
collectively as a society,
collectively as a society,
36
139661
5272
所以换句话说，
在作为集体的这个社会里，
在作为集体的这个社会里，
我们依然非常茫然，因为我们最智能的机器
依然有视觉上的缺陷。
依然有视觉上的缺陷。
02:36
we're very much blind,
37
144933
1746
02:38
because our smartest
machines are still blind.
machines are still blind.
38
146679
3387
02:43
"Why is this so hard?" you may ask.
39
151526
2926
”为什么这么困难？“你也许会问。
照相机可以像这样获得照片：
02:46
Cameras can take pictures like this one
40
154452
2693
它把采集到的光线转换成
二维数字矩阵来存储
二维数字矩阵来存储
02:49
by converting lights into
a two-dimensional array of numbers
a two-dimensional array of numbers
41
157145
3994
——也就是“像素”，
02:53
known as pixels,
42
161139
1650
但这些仍然是死板的数字。
02:54
but these are just lifeless numbers.
43
162789
2251
02:57
They do not carry meaning in themselves.
44
165040
3111
它们自身并不携带任何意义。
03:00
Just like to hear is not
the same as to listen,
the same as to listen,
45
168151
4343
就像”听到“和”听“完全不同，
03:04
to take pictures is not
the same as to see,
the same as to see,
46
172494
4040
”拍照“和”看“也完全不同。
03:08
and by seeing,
we really mean understanding.
we really mean understanding.
47
176534
3829
通过“看”，
我们实际上是“理解”了这个画面。
我们实际上是“理解”了这个画面。
03:13
In fact, it took Mother Nature
540 million years of hard work
540 million years of hard work
48
181293
6177
事实上，大自然经过了5亿4千万年的努力
03:19
to do this task,
49
187470
1973
才完成了这个工作，
03:21
and much of that effort
50
189443
1881
而这努力中更多的部分
03:23
went into developing the visual
processing apparatus of our brains,
processing apparatus of our brains,
51
191324
5271
是用在进化我们的大脑内
用于视觉处理的器官，
用于视觉处理的器官，
03:28
not the eyes themselves.
52
196595
2647
而不是眼睛本身。
03:31
So vision begins with the eyes,
53
199242
2747
所以"视觉”从眼睛采集信息开始，
03:33
but it truly takes place in the brain.
54
201989
3518
但大脑才是它真正呈现意义的地方。
03:38
So for 15 years now, starting
from my Ph.D. at Caltech
from my Ph.D. at Caltech
55
206287
5060
所以15年来，
从我进入加州理工学院攻读Ph.D.
从我进入加州理工学院攻读Ph.D.
03:43
and then leading Stanford's Vision Lab,
56
211347
2926
到后来领导
斯坦福大学的视觉实验室，
斯坦福大学的视觉实验室，
03:46
I've been working with my mentors,
collaborators and students
collaborators and students
57
214273
4396
我一直在和我的导师、
合作者和学生们一起
合作者和学生们一起
03:50
to teach computers to see.
58
218669
2889
教计算机如何去“看”。
我们的研究领域叫做
"计算机视觉与机器学习"。
"计算机视觉与机器学习"。
03:54
Our research field is called
computer vision and machine learning.
computer vision and machine learning.
59
222658
3294
03:57
It's part of the general field
of artificial intelligence.
of artificial intelligence.
60
225952
3878
这是AI（人工智能）领域的一个分支。
04:03
So ultimately, we want to teach
the machines to see just like we do:
the machines to see just like we do:
61
231000
5493
最终，我们希望能教会机器
像我们一样看见事物：
像我们一样看见事物：
04:08
naming objects, identifying people,
inferring 3D geometry of things,
inferring 3D geometry of things,
62
236493
5387
识别物品、辨别不同的人、
推断物体的立体形状、
推断物体的立体形状、
04:13
understanding relations, emotions,
actions and intentions.
actions and intentions.
63
241880
5688
理解事物的关联、
人的情绪、动作和意图。
人的情绪、动作和意图。
04:19
You and I weave together entire stories
of people, places and things
of people, places and things
64
247568
6153
像你我一样，只凝视一个画面一眼
就能理清整个故事中的人物、地点、事件。
就能理清整个故事中的人物、地点、事件。
04:25
the moment we lay our gaze on them.
65
253721
2164
04:28
The first step towards this goal
is to teach a computer to see objects,
is to teach a computer to see objects,
66
256955
5583
实现这一目标的第一步是
教计算机看到“对象”（物品），
教计算机看到“对象”（物品），
04:34
the building block of the visual world.
67
262538
3368
这是建造视觉世界的基石。
04:37
In its simplest terms,
imagine this teaching process
imagine this teaching process
68
265906
4434
在这个最简单的任务里，
想象一下这个教学过程：
想象一下这个教学过程：
04:42
as showing the computers
some training images
some training images
69
270340
2995
给计算机看一些特定物品的训练图片，
比如说猫，
比如说猫，
04:45
of a particular object, let's say cats,
70
273335
3321
04:48
and designing a model that learns
from these training images.
from these training images.
71
276656
4737
并让它从这些训练图片中，
学习建立出一个模型来。
学习建立出一个模型来。
04:53
How hard can this be?
72
281393
2044
这有多难呢？
不管怎么说，一只猫只是一些
形状和颜色拼凑起来的图案罢了，
形状和颜色拼凑起来的图案罢了，
04:55
After all, a cat is just
a collection of shapes and colors,
a collection of shapes and colors,
73
283437
4052
04:59
and this is what we did
in the early days of object modeling.
in the early days of object modeling.
74
287489
4086
比如这个就是我们
最初设计的抽象模型。
最初设计的抽象模型。
05:03
We'd tell the computer algorithm
in a mathematical language
in a mathematical language
75
291575
3622
我们用数学的语言，
告诉计算机这种算法：
告诉计算机这种算法：
05:07
that a cat has a round face,
a chubby body,
a chubby body,
76
295197
3343
“猫”有着圆脸、胖身子、
05:10
two pointy ears, and a long tail,
77
298540
2299
两个尖尖的耳朵，还有一条长尾巴，
05:12
and that looked all fine.
78
300839
1410
这（算法）看上去挺好的。
但如果遇到这样的猫呢？
05:14
But what about this cat?
79
302859
2113
（笑）
05:16
(Laughter)
80
304972
1091
它整个蜷缩起来了。
05:18
It's all curled up.
81
306063
1626
现在你不得不加入一些别的形状和视角
来描述这个物品模型。
来描述这个物品模型。
05:19
Now you have to add another shape
and viewpoint to the object model.
and viewpoint to the object model.
82
307689
4719
05:24
But what if cats are hidden?
83
312408
1715
但如果猫是藏起来的呢？
05:27
What about these silly cats?
84
315143
2219
再看看这些傻猫呢？
05:31
Now you get my point.
85
319112
2417
你现在知道了吧。
05:33
Even something as simple
as a household pet
as a household pet
86
321529
3367
即使那些事物简单到
只是一只家养的宠物，
只是一只家养的宠物，
05:36
can present an infinite number
of variations to the object model,
of variations to the object model,
87
324896
4504
都可以出呈现出无限种变化的外观模型，
而这还只是“一个”对象的模型。
05:41
and that's just one object.
88
329400
2233
05:44
So about eight years ago,
89
332573
2492
所以大概在8年前，
05:47
a very simple and profound observation
changed my thinking.
changed my thinking.
90
335065
5030
一个非常简单、有冲击力的
观察改变了我的想法。
观察改变了我的想法。
05:53
No one tells a child how to see,
91
341425
2685
没有人教过婴儿怎么“看”，
05:56
especially in the early years.
92
344110
2261
尤其是在他们还很小的时候。
05:58
They learn this through
real-world experiences and examples.
real-world experiences and examples.
93
346371
5000
他们是从真实世界的经验
和例子中学到这个的。
和例子中学到这个的。
06:03
If you consider a child's eyes
94
351371
2740
如果你把孩子的眼睛
都看作是生物照相机，
06:06
as a pair of biological cameras,
95
354111
2554
06:08
they take one picture
about every 200 milliseconds,
about every 200 milliseconds,
96
356665
4180
那他们每200毫秒就拍一张照。
06:12
the average time an eye movement is made.
97
360845
3134
——这是眼球转动一次的平均时间。
所以到3岁大的时候，一个孩子已经看过了
上亿张的真实世界照片。
上亿张的真实世界照片。
06:15
So by age three, a child would have seen
hundreds of millions of pictures
hundreds of millions of pictures
98
363979
5550
06:21
of the real world.
99
369529
1834
这种“训练照片”的数量是非常大的。
06:23
That's a lot of training examples.
100
371363
2280
所以，与其孤立地关注于
算法的优化、再优化，
算法的优化、再优化，
06:26
So instead of focusing solely
on better and better algorithms,
on better and better algorithms,
101
374383
5989
我的关注点放在了给算法
提供像那样的训练数据
提供像那样的训练数据
06:32
my insight was to give the algorithms
the kind of training data
the kind of training data
102
380372
5272
06:37
that a child was given through experiences
103
385644
3319
——那些，婴儿们从经验中获得的
质量和数量都极其惊人的训练照片。
质量和数量都极其惊人的训练照片。
06:40
in both quantity and quality.
104
388963
3878
06:44
Once we know this,
105
392841
1858
一旦我们知道了这个，
06:46
we knew we needed to collect a data set
106
394699
2971
我们就明白自己需要收集的数据集，
06:49
that has far more images
than we have ever had before,
than we have ever had before,
107
397670
4459
必须比我们曾有过的任何数据库都丰富
06:54
perhaps thousands of times more,
108
402129
2577
——可能要丰富数千倍。
06:56
and together with Professor
Kai Li at Princeton University,
Kai Li at Princeton University,
109
404706
4111
因此，通过与普林斯顿大学的
Kai Li教授合作，
Kai Li教授合作，
07:00
we launched the ImageNet project in 2007.
110
408817
4752
我们在2007年发起了
ImageNet（图片网络）计划。
ImageNet（图片网络）计划。
07:05
Luckily, we didn't have to mount
a camera on our head
a camera on our head
111
413569
3838
幸运的是，我们不必在自己脑子里
装上一台照相机，然后等它拍很多年。
装上一台照相机，然后等它拍很多年。
07:09
and wait for many years.
112
417407
1764
我们运用了互联网，
07:11
We went to the Internet,
113
419171
1463
07:12
the biggest treasure trove of pictures
that humans have ever created.
that humans have ever created.
114
420634
4436
这个由人类创造的
最大的图片宝库。
最大的图片宝库。
07:17
We downloaded nearly a billion images
115
425070
3041
我们下载了接近10亿张图片
07:20
and used crowdsourcing technology
like the Amazon Mechanical Turk platform
like the Amazon Mechanical Turk platform
116
428111
5880
并利用众包技术（利用互联网分配工作、发现创意或
解决技术问题），像“亚马逊土耳其机器人”这样的平台
解决技术问题），像“亚马逊土耳其机器人”这样的平台
07:25
to help us to label these images.
117
433991
2339
来帮我们标记这些图片。
07:28
At its peak, ImageNet was one of
the biggest employers
the biggest employers
118
436330
4900
在高峰期时，ImageNet是「亚马逊土耳其机器人」
这个平台上最大的雇主之一：
这个平台上最大的雇主之一：
07:33
of the Amazon Mechanical Turk workers:
119
441230
2996
来自世界上167个国家的
接近5万个工作者，在一起工作
接近5万个工作者，在一起工作
07:36
together, almost 50,000 workers
120
444226
3854
07:40
from 167 countries around the world
121
448080
4040
帮我们筛选、排序、标记了
接近10亿张备选照片。
接近10亿张备选照片。
07:44
helped us to clean, sort and label
122
452120
3947
07:48
nearly a billion candidate images.
123
456067
3575
这就是我们为这个计划投入的精力，
07:52
That was how much effort it took
124
460612
2653
07:55
to capture even a fraction
of the imagery
of the imagery
125
463265
3900
去捕捉，一个婴儿可能在他早期发育阶段
获取的”一小部分“图像。
获取的”一小部分“图像。
07:59
a child's mind takes in
in the early developmental years.
in the early developmental years.
126
467165
4171
事后我们再来看，这个利用大数据来训练
计算机算法的思路，也许现在看起来很普通，
计算机算法的思路，也许现在看起来很普通，
08:04
In hindsight, this idea of using big data
127
472148
3902
08:08
to train computer algorithms
may seem obvious now,
may seem obvious now,
128
476050
4550
但回到2007年时，它就不那么寻常了。
08:12
but back in 2007, it was not so obvious.
129
480600
4110
我们在这段旅程上孤独地前行了很久。
08:16
We were fairly alone on this journey
for quite a while.
for quite a while.
130
484710
3878
一些很友善的同事建议我
做一些更有用的事来获得终身教职，
做一些更有用的事来获得终身教职，
08:20
Some very friendly colleagues advised me
to do something more useful for my tenure,
to do something more useful for my tenure,
131
488588
5003
08:25
and we were constantly struggling
for research funding.
for research funding.
132
493591
4342
而且我们也不断地为项目的研究经费发愁。
08:29
Once, I even joked to my graduate students
133
497933
2485
有一次，我甚至对
我的研究生学生开玩笑说：
我的研究生学生开玩笑说：
08:32
that I would just reopen
my dry cleaner's shop to fund ImageNet.
my dry cleaner's shop to fund ImageNet.
134
500418
4063
我要重新回去开我的干洗店
来赚钱资助ImageNet了。
来赚钱资助ImageNet了。
08:36
After all, that's how I funded
my college years.
my college years.
135
504481
4761
——毕竟，我的大学时光
就是靠这个资助的。
就是靠这个资助的。
所以我们仍然在继续着。
08:41
So we carried on.
136
509242
1856
在2009年，ImageNet项目诞生了——
08:43
In 2009, the ImageNet project delivered
137
511098
3715
08:46
a database of 15 million images
138
514813
4042
一个含有1500万张照片的数据库，
涵盖了22000种物品。
涵盖了22000种物品。
08:50
across 22,000 classes
of objects and things
of objects and things
139
518855
4805
这些物品是根据日常英语单词
进行分类组织的。
进行分类组织的。
08:55
organized by everyday English words.
140
523660
3320
无论是在质量上还是数量上，
08:58
In both quantity and quality,
141
526980
2926
这都是一个规模空前的数据库。
09:01
this was an unprecedented scale.
142
529906
2972
09:04
As an example, in the case of cats,
143
532878
3461
举个例子，在"猫"这个对象中，
09:08
we have more than 62,000 cats
144
536339
2809
我们有超过62000只猫
09:11
of all kinds of looks and poses
145
539148
4110
长相各异，姿势五花八门，
09:15
and across all species
of domestic and wild cats.
of domestic and wild cats.
146
543258
5223
而且涵盖了各种品种的家猫和野猫。
我们对ImageNet收集到的图片
感到异常兴奋，
感到异常兴奋，
09:20
We were thrilled
to have put together ImageNet,
to have put together ImageNet,
147
548481
3344
09:23
and we wanted the whole research world
to benefit from it,
to benefit from it,
148
551825
3738
而且我们希望整个研究界能从中受益，
所以以一种和TED一样的方式，
09:27
so in the TED fashion,
we opened up the entire data set
we opened up the entire data set
149
555563
4041
我们公开了整个数据库，
免费提供给全世界的研究团体。
免费提供给全世界的研究团体。
09:31
to the worldwide
research community for free.
research community for free.
150
559604
3592
（掌声）
09:36
(Applause)
151
564636
4000
那么现在，我们有了用来
培育计算机大脑的数据库，
培育计算机大脑的数据库，
09:41
Now that we have the data
to nourish our computer brain,
to nourish our computer brain,
152
569416
4538
我们可以回到”算法“本身上来了。
09:45
we're ready to come back
to the algorithms themselves.
to the algorithms themselves.
153
573954
3737
因为ImageNet的横空出世，它提供的信息财富
完美地适用于一些特定类别的机器学习算法，
完美地适用于一些特定类别的机器学习算法，
09:49
As it turned out, the wealth
of information provided by ImageNet
of information provided by ImageNet
154
577691
5178
09:54
was a perfect match to a particular class
of machine learning algorithms
of machine learning algorithms
155
582869
4806
称作“卷积神经网络”，
09:59
called convolutional neural network,
156
587675
2415
10:02
pioneered by Kunihiko Fukushima,
Geoff Hinton, and Yann LeCun
Geoff Hinton, and Yann LeCun
157
590090
5248
最早由Kunihiko Fukushima，Geoff Hinton，
和Yann LeCun在上世纪七八十年代开创。
和Yann LeCun在上世纪七八十年代开创。
10:07
back in the 1970s and '80s.
158
595338
3645
就像大脑是由上十亿的
紧密联结的神经元组成，
紧密联结的神经元组成，
10:10
Just like the brain consists
of billions of highly connected neurons,
of billions of highly connected neurons,
159
598983
5619
10:16
a basic operating unit in a neural network
160
604602
3854
神经网络里最基础的运算单元
也是一个“神经元式”的节点。
也是一个“神经元式”的节点。
10:20
is a neuron-like node.
161
608456
2415
10:22
It takes input from other nodes
162
610871
2554
每个节点从其它节点处获取输入信息，
然后把自己的输出信息再交给另外的节点。
然后把自己的输出信息再交给另外的节点。
10:25
and sends output to others.
163
613425
2718
此外，这些成千上万、甚至上百万的节点
10:28
Moreover, these hundreds of thousands
or even millions of nodes
or even millions of nodes
164
616143
4713
10:32
are organized in hierarchical layers,
165
620856
3227
都被按等级分布于不同层次，
就像大脑一样。
10:36
also similar to the brain.
166
624083
2554
在一个我们用来训练“对象识别模型”的
典型神经网络里，
典型神经网络里，
10:38
In a typical neural network we use
to train our object recognition model,
to train our object recognition model,
167
626637
4783
10:43
it has 24 million nodes,
168
631420
3181
有着2400万个节点，1亿4千万个参数，
和150亿个联结。
和150亿个联结。
10:46
140 million parameters,
169
634601
3297
10:49
and 15 billion connections.
170
637898
2763
这是一个庞大的模型。
10:52
That's an enormous model.
171
640661
2415
10:55
Powered by the massive data from ImageNet
172
643076
3901
借助ImageNet提供的巨大规模数据支持，
通过大量最先进的CPU和GPU，
来训练这些堆积如山的模型，
来训练这些堆积如山的模型，
10:58
and the modern CPUs and GPUs
to train such a humongous model,
to train such a humongous model,
173
646977
5433
“卷积神经网络”
以难以想象的方式蓬勃发展起来。
以难以想象的方式蓬勃发展起来。
11:04
the convolutional neural network
174
652410
2369
11:06
blossomed in a way that no one expected.
175
654779
3436
它成为了一个成功体系，
11:10
It became the winning architecture
176
658215
2508
11:12
to generate exciting new results
in object recognition.
in object recognition.
177
660723
5340
在对象识别领域，
产生了激动人心的新成果。
产生了激动人心的新成果。
11:18
This is a computer telling us
178
666063
2810
这张图，是计算机在告诉我们：
11:20
this picture contains a cat
179
668873
2300
照片里有一只猫、
11:23
and where the cat is.
180
671173
1903
还有猫所在的位置。
当然不止有猫了，
11:25
Of course there are more things than cats,
181
673076
2112
所以这是计算机算法在告诉我们
11:27
so here's a computer algorithm telling us
182
675188
2438
照片里有一个男孩，和一个泰迪熊；
11:29
the picture contains
a boy and a teddy bear;
a boy and a teddy bear;
183
677626
3274
一只狗，一个人，和背景里的小风筝；
11:32
a dog, a person, and a small kite
in the background;
in the background;
184
680900
4366
或者是一张拍摄于闹市的照片
比如人、滑板、栏杆、灯柱…等等。
比如人、滑板、栏杆、灯柱…等等。
11:37
or a picture of very busy things
185
685266
3135
11:40
like a man, a skateboard,
railings, a lampost, and so on.
railings, a lampost, and so on.
186
688401
4644
有时候，如果计算机
不是很确定它看到的是什么，
不是很确定它看到的是什么，
11:45
Sometimes, when the computer
is not so confident about what it sees,
is not so confident about what it sees,
187
693045
5293
11:51
we have taught it to be smart enough
188
699498
2276
我们还教它用足够聪明的方式
给出一个“安全”的答案，而不是“言多必失”
给出一个“安全”的答案，而不是“言多必失”
11:53
to give us a safe answer
instead of committing too much,
instead of committing too much,
189
701774
3878
——就像人类面对这类问题时一样。
11:57
just like we would do,
190
705652
2811
但在其他时候，我们的计算机
算法厉害到可以告诉我们
算法厉害到可以告诉我们
12:00
but other times our computer algorithm
is remarkable at telling us
is remarkable at telling us
191
708463
4666
12:05
what exactly the objects are,
192
713129
2253
关于对象的更确切的信息，
比如汽车的品牌、型号、年份。
比如汽车的品牌、型号、年份。
12:07
like the make, model, year of the cars.
193
715382
3436
12:10
We applied this algorithm to millions
of Google Street View images
of Google Street View images
194
718818
5386
我们在上百万张谷歌街景照片中
应用了这一算法，
应用了这一算法，
12:16
across hundreds of American cities,
195
724204
3135
那些照片涵盖了上百个美国城市。
我们从中发现一些有趣的事：
12:19
and we have learned something
really interesting:
really interesting:
196
727339
2926
首先，它证实了我们的一些常识：
12:22
first, it confirmed our common wisdom
197
730265
3320
12:25
that car prices correlate very well
198
733585
3290
汽车的价格，与家庭收入
呈现出明显的正相关。
呈现出明显的正相关。
12:28
with household incomes.
199
736875
2345
但令人惊奇的是，汽车价格与犯罪率
也呈现出明显的正相关性，
也呈现出明显的正相关性，
12:31
But surprisingly, car prices
also correlate well
also correlate well
200
739220
4527
12:35
with crime rates in cities,
201
743747
2300
以上结论是基于城市、或投票的
邮编区域进行分析的结果。
邮编区域进行分析的结果。
12:39
or voting patterns by zip codes.
202
747007
3963
那么等一下，这就是全部成果了吗？
12:44
So wait a minute. Is that it?
203
752060
2206
计算机是不是已经达到，
或者甚至超过了人类的能力？
或者甚至超过了人类的能力？
12:46
Has the computer already matched
or even surpassed human capabilities?
or even surpassed human capabilities?
204
754266
5153
12:51
Not so fast.
205
759419
2138
——还没有那么快。
目前为止，我们还只是
教会了计算机去看对象。
教会了计算机去看对象。
12:53
So far, we have just taught
the computer to see objects.
the computer to see objects.
206
761557
4923
12:58
This is like a small child
learning to utter a few nouns.
learning to utter a few nouns.
207
766480
4644
这就像是一个小宝宝学会说出几个名词。
这是一项难以置信的成就，
13:03
It's an incredible accomplishment,
208
771124
2670
但这还只是第一步。
13:05
but it's only the first step.
209
773794
2460
很快，我们就会到达
发展历程的另一个里程碑：
发展历程的另一个里程碑：
13:08
Soon, another developmental
milestone will be hit,
milestone will be hit,
210
776254
3762
这个小孩会开始用“句子”进行交流。
13:12
and children begin
to communicate in sentences.
to communicate in sentences.
211
780016
3461
13:15
So instead of saying
this is a cat in the picture,
this is a cat in the picture,
212
783477
4224
所以不止是说这张图里有只“猫”，
13:19
you already heard the little girl
telling us this is a cat lying on a bed.
telling us this is a cat lying on a bed.
213
787701
5202
你在开头已经听到小妹妹
告诉我们“这只猫是坐在床上的”。
告诉我们“这只猫是坐在床上的”。
13:24
So to teach a computer
to see a picture and generate sentences,
to see a picture and generate sentences,
214
792903
5595
为了教计算机看懂图片并生成句子，
13:30
the marriage between big data
and machine learning algorithm
and machine learning algorithm
215
798498
3948
“大数据”和“机器学习算法”的结合
需要更进一步。
需要更进一步。
13:34
has to take another step.
216
802446
2275
现在，计算机需要从图片和人类创造的
自然语言句子中同时进行学习。
自然语言句子中同时进行学习。
13:36
Now, the computer has to learn
from both pictures
from both pictures
217
804721
4156
13:40
as well as natural language sentences
218
808877
2856
13:43
generated by humans.
219
811733
3322
就像我们的大脑，
把视觉现象和语言融合在一起，
把视觉现象和语言融合在一起，
13:47
Just like the brain integrates
vision and language,
vision and language,
220
815055
3853
13:50
we developed a model
that connects parts of visual things
that connects parts of visual things
221
818908
5201
我们开发了一个模型，
可以把一部分视觉信息，像视觉片段，
与语句中的文字、短语联系起来。
与语句中的文字、短语联系起来。
13:56
like visual snippets
222
824109
1904
13:58
with words and phrases in sentences.
223
826013
4203
14:02
About four months ago,
224
830216
2763
大约4个月前，
我们最终把所有技术结合在了一起，
我们最终把所有技术结合在了一起，
14:04
we finally tied all this together
225
832979
2647
创造了第一个“计算机视觉模型”，
14:07
and produced one of the first
computer vision models
computer vision models
226
835626
3784
14:11
that is capable of generating
a human-like sentence
a human-like sentence
227
839410
3994
它在看到图片的第一时间，就有能力生成
类似人类语言的句子。
类似人类语言的句子。
14:15
when it sees a picture for the first time.
228
843404
3506
现在，我准备给你们看看
计算机看到图片时会说些什么
计算机看到图片时会说些什么
14:18
Now, I'm ready to show you
what the computer says
what the computer says
229
846910
4644
14:23
when it sees the picture
230
851554
1975
——还是那些在演讲开头给小女孩看的图片。
14:25
that the little girl saw
at the beginning of this talk.
at the beginning of this talk.
231
853529
3830
（视频）计算机：
“一个男人站在一头大象旁边。”
“一个男人站在一头大象旁边。”
14:31
(Video) Computer: A man is standing
next to an elephant.
next to an elephant.
232
859519
3344
“一架大飞机停在机场跑道一端。”
14:36
A large airplane sitting on top
of an airport runway.
of an airport runway.
233
864393
3634
14:41
FFL: Of course, we're still working hard
to improve our algorithms,
to improve our algorithms,
234
869057
4212
李飞飞：
当然，我们还在努力改善我们的算法，
当然，我们还在努力改善我们的算法，
14:45
and it still has a lot to learn.
235
873269
2596
它还有很多要学的东西。
（掌声）
14:47
(Applause)
236
875865
2291
计算机还是会犯很多错误的。
14:51
And the computer still makes mistakes.
237
879556
3321
（视频）计算机：
“一只猫躺在床上的毯子上。”
“一只猫躺在床上的毯子上。”
14:54
(Video) Computer: A cat lying
on a bed in a blanket.
on a bed in a blanket.
238
882877
3391
李飞飞：所以…当然——如果它看过太多种的猫，
它就会觉得什么东西都长得像猫……
它就会觉得什么东西都长得像猫……
14:58
FFL: So of course, when it sees
too many cats,
too many cats,
239
886268
2553
15:00
it thinks everything
might look like a cat.
might look like a cat.
240
888821
2926
（视频）计算机：
“一个小男孩拿着一根棒球棍。”
“一个小男孩拿着一根棒球棍。”
15:05
(Video) Computer: A young boy
is holding a baseball bat.
is holding a baseball bat.
241
893317
2864
15:08
(Laughter)
242
896181
1765
（笑声）
李飞飞：或者…如果它从没见过牙刷，
它就分不清牙刷和棒球棍的区别。
它就分不清牙刷和棒球棍的区别。
15:09
FFL: Or, if it hasn't seen a toothbrush,
it confuses it with a baseball bat.
it confuses it with a baseball bat.
243
897946
4583
（视频）计算机：
“建筑旁的街道上有一个男人骑马经过。”
“建筑旁的街道上有一个男人骑马经过。”
15:15
(Video) Computer: A man riding a horse
down a street next to a building.
down a street next to a building.
244
903309
3434
15:18
(Laughter)
245
906743
2023
（笑声）
李飞飞：我们还没教它Art 101
（美国大学艺术基础课）。
（美国大学艺术基础课）。
15:20
FFL: We haven't taught Art 101
to the computers.
to the computers.
246
908766
3552
（视频）计算机：
“一只斑马站在一片草原上。”
“一只斑马站在一片草原上。”
15:25
(Video) Computer: A zebra standing
in a field of grass.
in a field of grass.
247
913768
2884
李飞飞：它还没学会像你我一样
欣赏大自然里的绝美景色。
欣赏大自然里的绝美景色。
15:28
FFL: And it hasn't learned to appreciate
the stunning beauty of nature
the stunning beauty of nature
248
916652
3367
15:32
like you and I do.
249
920019
2438
所以，这是一条漫长的道路。
15:34
So it has been a long journey.
250
922457
2832
将一个孩子从出生培养到3岁是很辛苦的。
15:37
To get from age zero to three was hard.
251
925289
4226
而真正的挑战是从3岁到13岁的过程中，
而且远远不止于此。
而且远远不止于此。
15:41
The real challenge is to go
from three to 13 and far beyond.
from three to 13 and far beyond.
252
929515
5596
让我再给你们看看这张
关于小男孩和蛋糕的图。
关于小男孩和蛋糕的图。
15:47
Let me remind you with this picture
of the boy and the cake again.
of the boy and the cake again.
253
935111
4365
目前为止，
我们已经教会计算机“看”对象，
我们已经教会计算机“看”对象，
15:51
So far, we have taught
the computer to see objects
the computer to see objects
254
939476
4064
15:55
or even tell us a simple story
when seeing a picture.
when seeing a picture.
255
943540
4458
或者甚至基于图片，
告诉我们一个简单的故事。
告诉我们一个简单的故事。
（视频）计算机：
”一个人坐在放蛋糕的桌子旁。“
”一个人坐在放蛋糕的桌子旁。“
15:59
(Video) Computer: A person sitting
at a table with a cake.
at a table with a cake.
256
947998
3576
李飞飞：但图片里还有更多信息
——远不止一个人和一个蛋糕。
——远不止一个人和一个蛋糕。
16:03
FFL: But there's so much more
to this picture
to this picture
257
951574
2630
16:06
than just a person and a cake.
258
954204
2270
计算机无法理解的是：
这是一个特殊的意大利蛋糕，
这是一个特殊的意大利蛋糕，
16:08
What the computer doesn't see
is that this is a special Italian cake
is that this is a special Italian cake
259
956474
4467
它只在复活节限时供应。
16:12
that's only served during Easter time.
260
960941
3217
而这个男孩穿着的
是他最喜欢的T恤衫，
是他最喜欢的T恤衫，
16:16
The boy is wearing his favorite t-shirt
261
964158
3205
那是他父亲去悉尼旅行时
带给他的礼物。
带给他的礼物。
16:19
given to him as a gift by his father
after a trip to Sydney,
after a trip to Sydney,
262
967363
3970
另外，你和我都能清楚地看出，
这个小孩有多高兴，以及这一刻在想什么。
这个小孩有多高兴，以及这一刻在想什么。
16:23
and you and I can all tell how happy he is
263
971333
3808
16:27
and what's exactly on his mind
at that moment.
at that moment.
264
975141
3203
这是我的儿子Leo。
16:31
This is my son Leo.
265
979214
3125
在我探索视觉智能的道路上，
16:34
On my quest for visual intelligence,
266
982339
2624
我不断地想到Leo
和他未来将要生活的那个世界。
和他未来将要生活的那个世界。
16:36
I think of Leo constantly
267
984963
2391
16:39
and the future world he will live in.
268
987354
2903
当机器可以“看到”的时候，
16:42
When machines can see,
269
990257
2021
医生和护士会获得一双额外的、
不知疲倦的眼睛，
不知疲倦的眼睛，
16:44
doctors and nurses will have
extra pairs of tireless eyes
extra pairs of tireless eyes
270
992278
4712
帮他们诊断病情、照顾病人。
16:48
to help them to diagnose
and take care of patients.
and take care of patients.
271
996990
4092
16:53
Cars will run smarter
and safer on the road.
and safer on the road.
272
1001082
4383
汽车可以在道路上行驶得
更智能、更安全。
更智能、更安全。
机器人，而不只是人类，
16:57
Robots, not just humans,
273
1005465
2694
会帮我们救助灾区被困和受伤的人员。
17:00
will help us to brave the disaster zones
to save the trapped and wounded.
to save the trapped and wounded.
274
1008159
4849
我们会发现新的物种、更好的材料，
17:05
We will discover new species,
better materials,
better materials,
275
1013798
3796
还可以在机器的帮助下
探索从未见到过的前沿地带。
探索从未见到过的前沿地带。
17:09
and explore unseen frontiers
with the help of the machines.
with the help of the machines.
276
1017594
4509
一点一点地，
我们正在赋予机器以视力。
我们正在赋予机器以视力。
17:15
Little by little, we're giving sight
to the machines.
to the machines.
277
1023113
4167
首先，我们教它们去“看”。
17:19
First, we teach them to see.
278
1027280
2798
然后，它们反过来也帮助我们，
让我们看得更清楚。
让我们看得更清楚。
17:22
Then, they help us to see better.
279
1030078
2763
17:24
For the first time, human eyes
won't be the only ones
won't be the only ones
280
1032841
4165
这是第一次，人类的眼睛不再
独自地思考和探索我们的世界。
独自地思考和探索我们的世界。
17:29
pondering and exploring our world.
281
1037006
2934
我们将不止是“使用”机器的智力，
17:31
We will not only use the machines
for their intelligence,
for their intelligence,
282
1039940
3460
我们还要以一种从未想象过的方式，
与它们“合作”。
与它们“合作”。
17:35
we will also collaborate with them
in ways that we cannot even imagine.
in ways that we cannot even imagine.
283
1043400
6179
我所追求的是：
17:41
This is my quest:
284
1049579
2161
赋予计算机视觉智能，
17:43
to give computers visual intelligence
285
1051740
2712
并为Leo和这个世界，
创造出更美好的未来。
创造出更美好的未来。
17:46
and to create a better future
for Leo and for the world.
for Leo and for the world.
286
1054452
5131
谢谢。
17:51
Thank you.
287
1059583
1811
（掌声）
17:53
(Applause)
288
1061394
3785
ABOUT THE SPEAKER
Fei-Fei Li - Computer scientistAs Director of Stanford’s Artificial Intelligence Lab and Vision Lab, Fei-Fei Li is working to solve AI’s trickiest problems -- including image recognition, learning and language processing.
Why you should listen
Using algorithms built on machine learning methods such as neural network models, the Stanford Artificial Intelligence Lab led by Fei-Fei Li has created software capable of recognizing scenes in still photographs -- and accurately describe them using natural language.
Li’s work with neural networks and computer vision (with Stanford’s Vision Lab) marks a significant step forward for AI research, and could lead to applications ranging from more intuitive image searches to robots able to make autonomous decisions in unfamiliar situations.
Fei-Fei was honored as one of Foreign Policy's 2015 Global Thinkers.
Fei-Fei Li | Speaker | TED.com