sponsored links
TED2003

Torsten Reil: Animate characters by evolving them

トーステン・レイル: 生物学を用いたアニメーション制作

March 3, 2003

トーステン・レイルは体の仕組みを研究し、骨・筋肉・神経システムを使って内側から人間を再現することにより、自然な動作をする人間のアニメーションを作ることができるようになると話します。このTED 2003でプレゼンでの成果はGTA4でご覧になれます。

Torsten Reil - Animating neurobiologist
By coding computer simulations with biologically modeled nervous systems, Torsten Reil and his company NaturalMotion breathe life into the animated characters inhabiting the most eye-poppingly realistic games and movies around. Full bio

sponsored links
Double-click the English subtitles below to play the video.
I'm going to talk about a technology that we're developing at Oxford now,
今日お話するのは我々が現在オックスフォードで
開発中の技術のことで
00:15
that we think is going to change the way that
コンピューターゲームやハリウッド映画の
00:19
computer games and Hollywood movies are being made.
制作法を変えると考えています
00:22
That technology is simulating humans.
その技術とは人間シミュレーションです
00:26
It's simulated humans with a simulated body
模擬人体とその体をコントロールする
00:29
and a simulated nervous system to control that body.
模擬神経システムを持った
シミュレートされた人間です
00:32
Now, before I talk more about that technology,
さて その技術についてお話する前に
00:36
let's have a quick look at what human characters look like
現在のゲームの人間のキャラクターは
00:39
at the moment in computer games.
どんなものなのかちょっと見てみましょう
00:42
This is a clip from a game called "Grand Theft Auto 3."
『グランド・セフト・オート3』というゲームのクリップです
00:45
We already saw that briefly yesterday.
昨日ちょっと見てみましたが
00:48
And what you can see is -- it is actually a very good game.
ご覧の通りとてもよく出来たゲームです
00:50
It's one of the most successful games of all time.
今までで最も成功したゲームの一つでしょう
00:53
But what you'll see is that all the animations in this game are very repetitive.
しかしお気付きの通り このゲームの
アニメーションは繰り返しばかりです
00:56
They pretty much look the same.
どれも 同じに見えます
01:00
I've made him run into a wall here, over and over again.
キャラクターを壁に
何度も何度もぶつけてみましたが
01:02
And you can see he looks always the same.
ご覧の通り常に同じ動きをします
01:05
The reason for that is that these characters
その理由はキャラクターが
01:07
are actually not real characters.
本物の人間ではないからです
01:10
They are a graphical visualization of a character.
これは人物を視覚化した
グラフィックでしかありません
01:12
To produce these animations, an animator at a studio has to anticipate
これらのアニメーションを作るために
スタジオのアニメーターは
01:16
what's going to happen in the actual game,
実際にゲーム中で何が起こるかを予測し
01:21
and then has to animate that particular sequence.
その一連の流れを
アニメーション化しなければなりません
01:24
So, he or she sits down, animates it, and tries to anticipate what's going to happen,
つまり机上でアニメーションを作る時
何が起こるのか予測する事で
01:27
and then these particular animations are just played back
このようなアニメーションがゲーム内の
01:31
at appropriate times in the computer game.
適切なタイミングで再生されるのです
01:34
Now, the result of that is that you can't have real interactivity.
この方法では真にインタラクティブな
ものにはなり得ず
01:37
All you have is animations that are played back
得られるのは妥当なタイミングで再生される
01:42
at more or less the appropriate times.
アニメーションに過ぎません
01:45
It also means that games aren't really going to be as surprising as they could be,
その結果 思ったように意外性のある
ゲームにはなり得ないのです
01:47
because you only get out of it, at least in terms of the character,
キャラクターに限っていえば
01:52
what you actually put into it.
組み込んだものしか出てこないのですから
01:55
There's no real emergence there.
何か新しい物が出てくるわけではありません
01:57
And thirdly, as I said, most of the animations are very repetitive because of that.
第三に 先述の通り アニメーション全体で
繰り返しが多いのも このためです
01:59
Now, the only way to get around that
これを解決する唯一の方法は
02:03
is to actually simulate the human body
実際に人間の体をシミュレーションし
02:05
and to simulate that bit of the nervous system of the brain that controls that body.
体をコントロールする脳神経系の一部を
シミュレートすることです
02:07
And maybe, if I could have you for a quick demonstration
違いを説明するために
02:12
to show what the difference is --
ちょっと実演してみます
02:15
because, I mean, it's very, very trivial.
見ればすぐにわかりますから
02:17
If I push Chris a bit, like this, for example, he'll react to it.
クリスをこんな感じにちょっと押すと
彼はそれに反応します
02:21
If I push him from a different angle, he'll react to it differently,
別の角度から押してみるとまた違った反応をします
02:24
and that's because he has a physical body,
それは彼に体が物理的に備わっていて
02:27
and because he has the motor skills to control that body.
その体をコントロールする運動機能を
持ち合わせているからです
02:29
It's a very trivial thing.
当たり前なことなんです
02:32
It's not something you get in computer games at the moment, at all.
現在まだこれはゲームに使われていません
02:34
Thank you very much. Chris Anderson: That's it?
ありがとう
クリス・アンダーソン: 終わり?
02:36
Torsten Reil: That's it, yes.
トーステン・レイル: はい 終わりです
02:38
So, that's what we're trying to simulate --
これを クリスだけに限らず人間全般で
シミュレートしようとしているわけです
02:40
not Chris specifically, I should say, but humans in general.
これを クリスだけに限らず人間全般で
シミュレートしようとしているわけです
02:41
Now, we started working on this a while ago at Oxford University,
この研究をオックスフォード大で
進めているわけですが
02:46
and we tried to start very simply.
まずは簡単なところから
02:51
What we tried to do was teach a stick figure how to walk.
棒人間に歩き方を教えることから始めました
02:53
That stick figure is physically stimulated. You can see it here on the screen.
この棒人間は物理的刺激を受けています
スクリーンでご覧下さい
02:56
So, it's subject to gravity, has joints, etc.
重力の影響を受けたり
関節があったりします
02:59
If you just run the simulation, it will just collapse, like this.
シミュレーションを行うだけだと
こんな感じに転んでしまいます
03:02
The tricky bit is now to put an AI controller in it
ここに人工知能の
制御装置を組み込んで
03:05
that actually makes it work.
これを上手く動かそうというわけです
03:09
And for that, we use the neural network, which we based on
そのために我々はニューラルネットワークを用います
03:11
that part of the nervous system that we have in our spine
背骨にある神経系の一部に基づいており
03:14
that controls walking in humans.
人間の歩行を制御します
03:16
It's called the central pattern generator.
中枢パターン発生器と呼ばれるものです
03:18
So, we simulated that as well, and then the really tricky bit
これもシミュレートしました
ここで 厄介なのは
03:20
is to teach that network how to walk.
このネットワークに歩行を教える事です
03:23
For that we used artificial evolution -- genetic algorithms.
そのために我々は人工的な進化を取り入れました
遺伝的アルゴリズムです
03:25
We heard about those already yesterday,
これについては昨日 話があったので
03:29
and I suppose that most of you are familiar with that already.
皆さんご存知かとは思いますが
03:31
But, just briefly, the concept is that
簡単に概念をご説明しますと
03:34
you create a large number of different individuals --
はじめに多数の互いに異なる個体 --
03:36
neural networks, in this case --
この場合はニューラルネットワークを --
03:39
all of which are random at the beginning.
ランダムに生成します
03:41
You hook these up -- in this case, to the virtual muscles
そしてこれらを仮想筋肉に繋げ--
03:43
of that two-legged creature here --
ここでは この2本足の動物のですが--
03:45
and hope that it does something interesting.
何か面白い事が起こるのを待つわけです
03:48
At the beginning, they're all going to be very boring.
最初はつまらない事ばかりです
03:51
Most of them won't move at all,
ほとんどは動きすらしないでしょう
03:53
but some of them might make a tiny step.
しかし動き出すものもあるかもしれません
03:55
Those are then selected by the algorithm,
そのような個体はアルゴリズムにより選ばれ
03:57
reproduced with mutation and recombinations to introduce sex as well.
突然変異と交叉を交えて複製し 交配もさせます
03:59
And you repeat that process over and over again,
このプロセスを 歩く個体が得られるまで
04:03
until you have something that walks --
何度も何度も繰り返します
04:05
in this case, in a straight line, like this.
この場合 このように真っ直ぐ歩くまでです
04:07
So that was the idea behind this.
これが背景となるアイデアです
04:09
When we started this, I set up the simulation one evening.
研究初期のある晩
シミュレーションを立ち上げ
04:11
It took about three to four hours to run the simulation.
3~4時間程 実行しました
04:14
I got up the next morning, went to the computer and looked at the results,
翌朝起きてコンピュータのところに
結果を見に行きました
04:17
and was hoping for something that walked in a straight line,
一つくらいは真っ直ぐに歩いていてくれ
と期待しながらです
04:21
like I've just demonstrated,
ちょうど今やってみたように
04:24
and this is what I got instead.
すると得られたのはこれでした
04:26
(Laughter)
(笑)
04:28
So, it was back to the drawing board for us.
ふりだしに戻ってしまいました
04:38
We did get it to work eventually,
試行錯誤を重ねて 最終的には
04:42
after tweaking a bit here and there.
歩かせることができました
04:45
And this is an example of a successful evolutionary run.
これは進化過程の成功例です
04:47
So, what you'll see in a moment is a very simple biped
これから単純な二足動物が 人工進化を用いて
04:50
that's learning how to walk using artificial evolution.
歩き方を学習していく様子をご覧頂きましょう
04:53
At the beginning, it can't walk at all,
最初は全く歩くことはできませんが
04:56
but it will get better and better over time.
時間とともに改善されていきます
04:58
So, this is the one that can't walk at all.
これは全く歩けないものです
05:02
(Laughter)
(笑)
05:05
Now, after five generations of applying evolutionary process,
5世代分の進化過程を適用すると
05:11
the genetic algorithm is getting a tiny bit better.
遺伝的アルゴリズムはほんの少しだけ良くなっていきます
05:14
(Laughter)
(笑)
05:17
Generation 10 and it'll take a few steps more --
第10世代では幾分かはマシになりますが
05:25
still not quite there.
歩けるようになったわけではありません
05:31
But now, after generation 20, it actually walks in a straight line without falling over.
しかし 第20世代を過ぎた段階では
転ばずに真っすぐ歩くようになります
05:34
That was the real breakthrough for us.
我々にとって大きな進歩でした
05:40
It was, academically, quite a challenging project,
学問的に極めて難しいプロジェクトでしたが
05:43
and once we had reached that stage, we were quite confident
この段階まで来ると他のことにも
05:46
that we could try and do other things as well with this approach --
このアプローチが使えるだろうと確信できました
05:49
actually simulating the body
人体をシミュレートするのに
05:52
and simulating that part of the nervous system that controls it.
神経系の必要な部分を
シミュレートできるはずです
05:54
Now, at this stage, it also became clear that this could be very exciting
ここまで来ると ゲームやネット上の世界で
05:57
for things like computer games or online worlds.
どう使えるか楽しみにもなります
06:00
What you see here is the character standing there,
これはキャラクターが立っていて
06:03
and there's an obstacle that we put in its way.
その行く先に障害物を配置した様子です
06:05
And what you see is, it's going to fall over the obstacle.
このキャラクターが障害物に
つまずく様子をご覧頂きましょう
06:07
Now, the interesting bit is, if I move the obstacle a tiny bit to the right,
面白いことに障害物を右にずらすと
06:12
which is what I'm doing now, here,
ここでやってみますが
06:15
it will fall over it in a completely different way.
全く異なる転び方をします
06:17
And again, if you move the obstacle a tiny bit, it'll again fall differently.
再び障害物をずらしてみると
また違ったように転びます
06:24
(Laughter)
(笑)
06:29
Now, what you see, by the way, at the top there,
上部に表示されているのは
06:31
are some of the neural activations being fed into the virtual muscles.
神経の活性化の一部が仮想筋肉に
送られている様子です
06:33
Okay. That's the video. Thanks.
以上ビデオでした
ありがとうございました
06:36
Now, this might look kind of trivial, but it's actually very important
当然のようでもこれが重要なのは
06:38
because this is not something you get at the moment
インタラクティブなものや仮想世界で
06:41
in any interactive or any virtual worlds.
まだ実用化されていないからです
06:43
Now, at this stage, we decided to start a company and move this further,
この段階で我々は会社を設立し
さらに先へ進むことにしました
06:48
because obviously this was just a very simple, blocky biped.
ごく簡単な二足動物で始めたものを
06:51
What we really wanted was a full human body.
人間の体全体でしたかったのです
06:54
So we started the company.
我々は会社を立ち上げ
06:56
We hired a team of physicists, software engineers and biologists
物理学者 ソフトウェアエンジニア
生物学者を雇ってチームを結成し
06:57
to work on this, and the first thing we had to work on
これに取り組みました
最初にしなければならなかったのは
07:02
was to create the human body, basically.
基本的には人間の体を作ることです
07:05
It's got to be relatively fast, so you can run it on a normal machine,
通常のコンピューター上で使えるよう
処理スピードが重要ですが
07:09
but it's got to be accurate enough, so it looks good enough, basically.
見栄えを考えると 精度も大切です
07:12
So we put quite a bit of biomechanical knowledge into this thing,
我々は 生体工学の知識を注ぎ込み
07:15
and tried to make it as realistic as possible.
できるだけ現実に近付けようと試みました
07:18
What you see here on the screen right now
今スクリーンでご覧頂いているのは
07:22
is a very simple visualization of that body.
その体をシンプルに視覚化したものです
07:24
I should add that it's very simple to add things like hair, clothes, etc.,
ちなみに髪や服などを
付け足すことも容易ですが
07:26
but what we've done here is use a very simple visualization,
ここではシンプルなモデルを使い
07:30
so you can concentrate on the movement.
動作に集中できるようにしました
07:33
Now, what I'm going to do right now, in a moment,
では このキャラクターをちょっと押して
07:35
is just push this character a tiny bit and we'll see what happens.
何が起こるのか見てみましょう
07:38
Nothing really interesting, basically.
とりたてて面白いことは起こりません
07:46
It falls over, but it falls over like a rag doll, basically.
転び方も まるでぬいぐるみの人形です
07:48
The reason for that is that there's no intelligence in it.
その理由は知能を持っていないからです
07:51
It becomes interesting when you put artificial intelligence into it.
人工知能を組み込むと面白くなります
07:54
So, this character now has motor skills in the upper body --
このキャラクターは上半身の運動スキルを
身に付けています
07:58
nothing in the legs yet, in this particular one.
この段階では脚にはまだですが
08:02
But what it will do -- I'm going to push it again.
もう一度押してみましょう このキャラクターは
08:04
It will realize autonomously that it's being pushed.
押されていることを自律的に認識し
08:07
It's going to stick out its hands.
手を突き出します
08:09
It's going to turn around into the fall, and try and catch the fall.
転ぶ方に体をひねって
地面に手をつこうとします
08:11
So that's what you see here.
ご覧の通り
08:20
Now, it gets really interesting
下半身にもAIを組み込んでみると
08:22
if you then add the AI for the lower part of the body as well.
更に面白いことが起こります
08:24
So here, we've got the same character.
ここでは同じキャラクターを使いますが
08:28
I'm going to push it a bit harder now,
前よりも少し強く押してみます
08:30
harder than I just pushed Chris.
クリスを押したのよりも強く
08:32
But what you'll see is -- it's going to receive a push now from the left.
このキャラクターは画面向かって左から押されます
08:34
What you see is it takes steps backwards,
数歩後ずさりながら
08:41
it tries to counter-balance,
バランスを取ろうとして
08:43
it tries to look at the place where it thinks it's going to land.
着地するであろう場所を見たのが
お分かり頂けたでしょう
08:45
I'll show you this again.
ではもう一度
08:49
And then, finally hits the floor.
最終的には地面に倒れます
08:51
Now, this becomes really exciting
私が今やってみせたように
このキャラクターを
08:55
when you push that character in different directions, again, just as I've done.
別の方向に押してみると
さらに面白くなります
08:58
That's something that you cannot do right now.
現在 このような事はできません
09:03
At the moment, you only have empty computer graphics in games.
今ゲームで使われているのは
空っぽのCG だけです
09:07
What this is now is a real simulation. That's what I want to show you now.
ここに実際のシミュレーションが
あるのでお見せします
09:10
So, here's the same character with the same behavior I've just shown you,
先ほどお見せしたのと同じキャラクター
同じ動作ですが
09:13
but now I'm just going to push it from different directions.
今度は別の方向から押してみましょう
09:16
First, starting with a push from the right.
まずは右から
09:18
This is all slow motion, by the way, so we can see what's going on.
ちなみに何が起こっているのか見れるように
全てスローモーションとなっています
09:23
Now, the angle will have changed a tiny bit,
では角度を少し変えてみると
09:26
so you can see that the reaction is different.
違う反応をするのがお分かり頂けるでしょう
09:29
Again, a push, now this time from the front.
もう一度 今度は正面から押してみましょう
09:33
And you see it falls differently.
そうすると また違う転び方をします
09:37
And now from the left --
では左から
09:39
and it falls differently.
するとまた違う転び方
09:43
That was really exciting for us to see that.
我々はこれを見て感動しました
09:45
That was the first time we've seen that.
こんなの見たことはなかったからです
09:47
This is the first time the public sees this as well,
これを今回初めて皆さんにも見ていただきます
09:49
because we have been in stealth mode.
ずっとステルスモードでやってきましたので
09:51
I haven't shown this to anybody yet.
まだ誰にも見せていないのです
09:53
Now, just a fun thing:
面白いものがあります
09:55
what happens if you put that character --
このキャラクターを--
09:57
this is now a wooden version of it, but it's got the same AI in it --
木製バージョンですが同じAIを持っています--
09:59
but if you put that character on a slippery surface, like ice.
氷のようなすべりやすい表面に置いたらどうなるでしょう
10:01
We just did that for a laugh, just to see what happens.
ネタとしてやってみました
10:03
(Laughter)
(笑)
10:06
And this is what happens.
こうなりました
10:07
(Laughter)
(笑)
10:09
(Applause)
(拍手)
10:12
It's nothing we had to do about this.
特別に何かをする必要は無く
10:15
We just took this character that I just talked about,
先ほどのキャラクターを
10:17
put it on a slippery surface, and this is what you get out of it.
つるつるした所に置くと
自然とこうなるのです
10:19
And that's a really fascinating thing about this approach.
これがこのアプローチの素晴らしさです
10:22
Now, when we went to film studios and games developers
映画スタジオやゲーム会社に行って
10:26
and showed them that technology, we got a very good response.
この技術を見せたところ
とても良い反応が得られました
10:29
And what they said was, the first thing they need immediately is virtual stuntmen.
彼らによると 今すぐ必要なのは
仮想スタントマンだそうです
10:32
Because stunts are obviously very dangerous, they're very expensive,
スタントは非常に危険なので
多額の費用がかかる上
10:36
and there are a lot of stunt scenes that you cannot do, obviously,
実際には行えないスタントシーンも多数あります
10:40
because you can't really allow the stuntman to be seriously hurt.
スタントマンに大怪我をさせるわけにはいきません
10:42
So, they wanted to have a digital version of a stuntman
ですので仮想スタントマンが
必要とされているのです
10:45
and that's what we've been working on for the past few months.
ここ数ヶ月それに取り組んできました
10:48
And that's our first product that we're going to release in a couple of weeks.
これが初作品で 数週間のうちにリリース予定です
10:50
So, here are just a few very simple scenes of the guy just being kicked.
男がただ蹴られるだけの
単純なシーンを用意しました
10:55
That's what people want. That's what we're giving them.
依頼されたとおりに 作っただけです
11:00
(Laughter)
(笑)
11:02
You can see, it's always reacting.
ご覧の通り常に反応しています
11:09
This is not a dead body. This is a body who basically, in this particular case,
まるで生きているかのように
この体はこの場合力を感じ取り
11:11
feels the force and tries to protect its head.
頭を守ろうとします
11:15
Only, I think it's quite a big blow again.
とても大きな衝撃だろうと思います
11:17
You feel kind of sorry for that thing,
皆さんは同情するかもしれませんが
11:19
and we've seen it so many times now that
我々は何度も繰り返して来たので
11:21
we don't really care any more.
誰も気にしなくなりました
11:23
(Laughter)
(笑)
11:25
There are much worse videos than this, by the way, which I have taken out, but ...
もっとひどいビデオもあるんですが
ここではお見せしません
11:26
Now, here's another one.
これは別の例
11:31
What people wanted as a behavior was to have an explosion,
依頼された動作は
爆発により
11:33
a strong force applied to the character,
大きな力がキャラクターにかかったときの
11:37
and have the character react to it in midair.
空中でキャラクターがどう反応するかです
11:39
So that you don't have a character that looks limp,
必要なのは ぐにゃぐにゃのキャラクターではなく
11:41
but actually a character that you can use in an action film straight away,
空中ではまるで生きているかのように見え
11:43
that looks kind of alive in midair as well.
アクション映画でそのまま使えるキャラクターです
11:46
So this character is going to be hit by a force,
このキャラクターが吹き飛ばされると
11:48
it's going to realize it's in the air,
空中にいるのだと認識し
11:50
and it's going to try and, well,
着地する方向へ
11:52
stick out its arm in the direction where it's landing.
手を突き出そうとします
11:55
That's one angle; here's another angle.
さっきとは違う角度から
11:59
We now think that the realism we're achieving with this
これだけリアルであれば
12:02
is good enough to be used in films.
映画でも十分使えます
12:04
And let's just have a look at a slightly different visualization.
少し違うのも見てみましょう
12:06
This is something I just got last night
我々のソフトウェアを試験的に使っている
12:09
from an animation studio in London, who are using our software
ロンドンのアニメスタジオから
12:11
and experimenting with it right now.
昨日届いたばかりです
12:14
So this is exactly the same behavior that you saw,
先ほど見たものと全く同じ動きですが
12:16
but in a slightly better rendered version.
より高品質にレンダリングされています
12:19
So if you look at the character carefully,
キャラクターをよく見てみると
12:23
you see there are lots of body movements going on,
体のあちこちが動いていますが
12:26
none of which you have to animate like in the old days.
これを動画化する必要はありません
12:28
Animators had to actually animate them.
従来はアニメーターが動画化したものです
12:30
This is all happening automatically in the simulation.
全てシミュレーションにより自動的に作られます
12:32
This is a slightly different angle,
では少し違う角度から
12:34
and again a slow motion version of this.
もう一度スローモーションで見てみましょう
12:39
This is incredibly quick. This is happening in real time.
とても高速です リアルタイムに
シミュレートされています
12:41
You can run this simulation in real time, in front of your eyes,
目の前でリアルタイムでシミュレーションを走らせ
12:45
change it, if you want to, and you get the animation straight out of it.
必要に応じてその場で修正することもできます
12:47
At the moment, doing something like this by hand
そのような作業は手作業だと
12:50
would take you probably a couple of days.
数日かかってしまうでしょう
12:52
This is another behavior they requested.
これはリクエストを受けたもう一つの動作です
12:55
I'm not quite sure why, but we've done it anyway.
用途は不明ですが ともかく作ってみました
12:58
It's a very simple behavior that shows you the power of this approach.
このアプローチの有用性を証明する
単純な動作です
13:00
In this case, the character's hands
この場合 キャラクターの手は
13:02
are fixed to a particular point in space,
空間中のある点に固定されていて
13:04
and all we've told the character to do is to struggle.
キャラクターに命令したのは もがくことだけです
13:06
And it looks organic. It looks realistic.
自然かつリアルに見えます
13:09
You feel kind of sorry for the guy.
可哀想にさえ感じてしまいます
13:12
It's even worse -- and that is another video I just got last night --
これはもっとひどいもの
昨夜届いたばかりのビデオですが__
13:14
if you render that a bit more realistically.
よりリアルにレンダリングされています
13:17
Now, I'm showing this to you just to show you
これをお見せしたのは
13:23
how organic it actually can feel, how realistic it can look.
自然さやリアルさを体感して頂くためです
13:25
And this is all a physical simulation of the body,
全身を物理シミュレーションで再現し
13:27
using AI to drive virtual muscles in that body.
AIで筋肉を仮想的に動かしました
13:30
Now, one thing which we did for a laugh was
ここで1つネタとして
13:35
to create a slightly more complex stunt scene,
少し複雑なスタントシーンを作ってみました
13:38
and one of the most famous stunts is the one where James Bond
ジェームス・ボンドがスイスのダムへ飛び込み
13:40
jumps off a dam in Switzerland and then is caught by a bungee.
バンジージャンプをする 有名なシーンです
13:43
Got a very short clip here.
ごく短い動画ですが
13:48
Yes, you can just about see it here.
ご覧頂きましょう
13:54
In this case, they were using a real stunt man. It was a very dangerous stunt.
映画では実際のスタントマンが使われました
とても危険なスタントでした
13:56
It was just voted, I think in the Sunday Times, as one of the most impressive stunts.
Sunday Timesだったと思いますが
最も印象的なスタントに選ばれました
13:59
Now, we've just tried and -- looked at our character and asked ourselves,
我々の作ったキャラクターを見て思いました
14:02
"Can we do that ourselves as well?"
「我々にもできるだろうか?」とね
14:05
Can we use the physical simulation of the character,
キャラクターの物理シミュレーションと
14:07
use artificial intelligence,
人工知能を用いて
14:09
put that artificial intelligence into the character,
その人工知能をキャラクターに組み込み
14:11
drive virtual muscles, simulate the way he jumps off the dam,
仮想的な筋肉を動かし
ダムに飛び込む様をシミュレートして
14:13
and then skydive afterwards,
そのあとスカイダイビングをさせた上で
14:17
and have him caught by a bungee afterwards?
バンジーのコードに引っかからせるのです
14:19
We did that. It took about altogether just two hours,
やってみました シミュレートするのに
14:21
pretty much, to create the simulation.
全体で2時間程しか かかりませんでした
14:24
And that's what it looks like, here.
こうなりました どうぞ
14:26
Now, this could do with a bit more work. It's still very early stages,
まだまだ改善の余地があります
これはまだ初期段階に過ぎません
14:37
and we pretty much just did this for a laugh,
単にできるかどうかを確認するために
14:40
just to see what we'd get out of it.
あくまでネタとして用意しただけですしね
14:42
But what we found over the past few months
しかしここ数ヶ月で分かったのは
14:44
is that this approach -- that we're pretty much standard upon --
我々が標準として用いているこのアプローチは
14:46
is incredibly powerful.
信じられないほど有用だということです
14:49
We are ourselves surprised what you actually get out of the simulations.
シミュレーションの出来に自分でも驚いています
14:51
There's very often very surprising behavior that you didn't predict before.
全く予想しなかった動作をすることも珍しくありません
14:55
There's so many things we can do with this right now.
他にも可能性は多数あります
14:59
The first thing, as I said, is going to be virtual stuntmen.
1つは先ほどお話した仮想スタントマンです
15:01
Several studios are using this software now to produce virtual stuntmen,
現在このソフトウェアを使って
仮想スタントマンを作っているスタジオもあり
15:04
and they're going to hit the screen quite soon, actually,
これを使ったメジャーな映画の作品が
15:08
for some major productions.
幾つか間もなく公開される予定です
15:10
The second thing is video games.
2つ目はテレビゲームです
15:12
With this technology, video games will look different and they will feel very different.
この技術を用いればテレビゲームはガラッと変わります
15:15
For the first time, you'll have actors that really feel very interactive,
真にインタラクティブに反応するキャラクターが
15:19
that have real bodies that really react.
初めてゲームに登場するわけです
15:22
I think that's going to be incredibly exciting.
本当に面白くなると思います
15:24
Probably starting with sports games,
おそらく始めはスポーツゲームで
15:27
which are going to become much more interactive.
よりインタラクティブになることでしょう
15:29
But I particularly am really excited
しかし個人的には
15:31
about using this technology in online worlds,
例えばトム・メルヒャーがやってくれたように
15:32
like there, for example, that Tom Melcher has shown us.
この技術をオンラインの世界で使うと
面白いのではないかと思います
15:35
The degree of interactivity you're going to get
インタラクティブ性は現在よりも
15:38
is totally different, I think, from what you're getting right now.
格段に向上するのではないかと思います
15:40
A third thing we are looking at and very interested in is simulation.
3番目に我々が興味を持っているのはシミュレーションです
15:44
We've been approached by several simulation companies,
シミュレーション会社 数社に
アプローチされましたが
15:49
but one project we're particularly excited about, which we're starting next month,
中でも特に楽しみなのが
来月開始予定のプロジェクトで
15:51
is to use our technology -- and in particular, the walking technology --
我々の技術 特に歩行技術を使用して
15:54
to help aid surgeons who work on children with cerebral palsy,
脳性小児麻痺の術後経過を予測することで
15:58
to predict the outcome of operations on these children.
小児外科医の補助をしようというものです
16:02
As you probably know,
恐らく皆さんご存知でしょうが
16:05
it's very difficult to predict what the outcome of an operation is
うまく歩けるように治療を施してみても
16:07
if you try and correct the gait.
手術の結果を予測するのは大変困難です
16:10
The classic quote is, I think, it's unpredictable at best,
古くからの格言にもこうあります
いくら頑張っても予測は不可能
16:12
is what people think right now, is the outcome.
結果について皆こう考えています
16:15
Now, what we want to do with our software is allow our surgeons to have a tool.
我々のソフトウエアを使って
外科医のツールを作りたいのです
16:18
We're going to simulate the gait of a particular child
我々の作る 特定の子供の歩行の
シミュレーションを使って
16:22
and the surgeon can then work on that simulation
医者は実際に手術を行う前に
16:25
and try out different ways to improve that gait,
歩行を改善する様々な方法を
試す事ができます
16:28
before he actually commits to an actual surgery.
歩行を改善する様々な方法を
試す事ができます
16:30
That's one project we're particularly excited about,
このプロジェクトはとても楽しみで
16:33
and that's going to start next month.
来月開始する予定です
16:35
Just finally, this is only just the beginning.
最後に申し上げますが
これは初めの一歩でしかありません
16:39
We can only do several behaviors right now.
まだ 行える動作に限りがあります
16:42
The AI isn't good enough to simulate a full human body.
人体を完全にシミュレートするには
AI はまだ不十分なのです
16:44
The body yes, but not all the motor skills that we have.
体については問題無いのですが
全ての運動技能の再生は無理なのです
16:47
And, I think, we're only there if we can have something like ballet dancing.
バレエのような動きができて初めて
うまくいったと言えると思います
16:50
Right now, we don't have that
現在はまだ そこまで行っていませんが
16:53
but I'm very sure that we will be able to do that at some stage.
いつかはそれが可能になると確信しています
16:55
We do have one unintentional dancer actually,
実は偶然生まれたダンサーを
16:57
the last thing I'm going to show you.
最後にお見せします
17:00
This was an AI contour that was produced and evolved --
AIにより生成された輪郭で 進化により —
17:02
half-evolved, I should say -- to produce balance, basically.
厳密には半進化と言うべきですが —
バランスをとるように 作られました
17:05
So, you kick the guy and the guy's supposed to counter-balance.
ですので キックすると逆方向にバランスを保とうとします
17:08
That's what we thought was going to come out of this.
そうなるはずと思っていましたが
17:11
But this is what emerged out of it, in the end.
結局得られたのはこれでした
17:14
(Music)
(音楽)
17:17
Bizarrely, this thing doesn't have a head. I'm not quite sure why.
なぜか頭はついていません
17:27
So, this was not something we actually put in there.
これは我々が組み込んだものではなく
17:31
He just started to create that dance himself.
このキャラクターが勝手に作ったダンスです
17:33
He's actually a better dancer than I am, I have to say.
ちなみに 私よりダンスは上手いようです
17:37
And what you see after a while --
しばらく見ていると
17:41
I think he even goes into a climax right at the end.
最後にクライマックスまで演出しています
17:43
And I think -- there you go.
おそらく — 始まりますよ
17:49
(Laughter)
(笑)
17:52
So, that all happened automatically. We didn't put that in there.
全く自動的に発生したもので
17:54
That's just the simulation creating this itself, basically.
我々ではなくシミュレーションが
ダンスを創作したのです
17:56
So it's just --
ですので —
17:59
(Applause)
(拍手)
18:01
Thanks.
ありがとう
18:02
Not quite John Travolta yet, but we're working on that as well,
まだジョン・トラボルタとは行きませんが
それにも取り組んでいます
18:05
so thanks very much for your time.
ご清聴ありがとうございました
18:08
Thanks.
ありがとう
18:10
(Applause)
(拍手)
18:11
CA: Incredible. That was really incredible.
CA: 素晴らしい 実に素晴らしい
18:12
TR: Thanks.
TR: ありがとう
18:14
Translator:Tomoshige Ohno
Reviewer:Akiko Hicks

sponsored links

Torsten Reil - Animating neurobiologist
By coding computer simulations with biologically modeled nervous systems, Torsten Reil and his company NaturalMotion breathe life into the animated characters inhabiting the most eye-poppingly realistic games and movies around.

Why you should listen
From modeling the mayhem of equine combat in Lord of the Rings: Return of the King to animating Liberty City gun battles in Grand Theft Auto IV, Torsten Reil's achievements are all over the map these days. Software that he helped create (with NaturalMotion, the imaging company he co-founded) has revolutionized computer animation of human and animal avatars, giving rise to some of the most breathtakingly real sequences in the virtual world of video games and movies- and along the way given valuable insight into the way human beings move their bodies.

Reil was a neural researcher working on his Masters at Oxford, developing computer simulations of nervous systems based on genetic algorithms-  programs that actually used natural selection to evolve their own means of locomotion. It didn't take long until he realized the commercial potential of these lifelike characters. In 2001 he capitalized on this lucrative adjunct to his research, and cofounded NaturalMotion. Since then the company has produced motion simulation programs like Euphoria and Morpheme, state of the art packages designed to drastically cut the time and expense of game development, and create animated worlds as real as the one outside your front door. Animation and special effects created with Endorphin (NaturalMotion's first animation toolkit) have lent explosive action to films such as Troy and Poseidon, and NaturalMotion's software is also being used by LucasArts in video games such as the hotly anticipated Indiana Jones.

But there are serious applications aside from the big screen and the XBox console: NaturalMotion has also worked under a grant from the British government to study the motion of a cerebral palsy patient, in hopes of finding therapies and surgeries that dovetail with the way her nervous system is functioning.
sponsored links

If you need translations, you can install "Google Translate" extension into your Chrome Browser.
Furthermore, you can change playback rate by installing "Video Speed Controller" extension.

Data provided by TED.

This website is owned and operated by Tokyo English Network.
The developer's blog is here.