like behind the scenes?

就像是在螢幕後面的嗎？

JOSS: Hmm, I have no idea.

JOSS: 恩恩，我也不知道。

JOE: Well do you think you can find out?

JOE: 呃～你覺得你可以找的出來（運作方式）嗎？

JOSS: You got it!

JOSS: 你找對人了！

These are what Snapchat calls their lenses, but everyone else calls filters.

Snapchat 把這稱為鏡頭，但是其他人都稱之為濾鏡。

They are very silly but the engineering behind them is serious.

雖然他們（的動作）很愚蠢，但是在這之後的工程是很認真的（嚴肅的）。

JOSS: Oh my god.

JOSS: 我的天呀！

The technology came from a Ukrainian startup called Looksery

這技術起於一位烏克蘭新創公司Looksery

which Snapchat acquired in September 2015 for a $150 million dollars.

Snapchat在2015年9月以1億5千萬元收購。

That's reportedly the largest tech acquisition in Ukrainian history.

據說是烏克蘭史上最大的一宗技術收購案。

Their augmented reality filters tap into the large and rapidly

他們增加了不少真實濾鏡，並貼在既大且快速成長的

growing field of "computer vision" --

領域「電腦影像」

those are applications that use pixel data from a camera in order to identify objects

這些用途皆使用由相機來的像素資料，來辨識物件

and interpret 3D space. Computer vision is how you can deposit checks,

並轉成3D空間。電腦影像是你為何可以付支票的原因，

it's how Facebook knows who's in your photos, how self-driving cars can avoid

讓電腦知道是誰在你的照片裡、自動駕駛如何能避開人群移動

running over people and how you can give yourself a doggy nose.

以及你怎麼給自己的一個小狗鼻子。

So how do snapchat filters work? They wouldn't let us talk to any of the Looksery

所以snapchat的濾鏡到底是怎麼運作的？他不會讓我們討問任何關於Looksery

engineers but their patents are online.

工程師，但他們的爸媽在線上。

The first step is detection. How does the computer know which part of an image is

第一步驟是偵查。電腦是怎麼知道影像中的哪一個部分是

a face?

臉呢？

This is something that human brains are fantastic at. Too good even.

這就是人類大腦中令人驚奇的一部分了。不能更好。

But this is what a photo looks like to a computer. If all you have is the data for

但這就是一張照片在電腦眼中看起來的樣子。如果你全部有的只是資訊

the color value of each individual pixel, how do you find a face?

每一個像素的色值，你要怎麼找到一張臉？

Well the key is looking for areas of contrast, between light and dark parts of

這關鍵就在於找到對比的區域，在光與影之間的影像。

the image. The pioneering facial detection tool is called the

先進的臉部偵查工具叫做

Viola-Jones algorithm.

Viola-Jones運算法。

It works by repeatedly scanning through the image data calculating the

他藉由重複掃描這些影片資料去計算

difference between the grayscale pixel values underneath the white boxes and

在白色與黑色區塊下的灰階像素之間的差異。

the black boxes. For instance, the bridge of the nose is usually lighter than the surrounding area on both sides,

例如，鼻樑通常比周圍還要亮

the eye sockets are darker than the forehead, and the middle of the forehead

眼窩比前額的顏色較深，額頭正中

is lighter than the size of it.

比眼窩大小的區塊還要亮。

These are crude test for facial features, but if they find enough matches in one

這些是對臉部特徵的粗糙測試，但如果他們在圖片中的一個區域找到足夠多的契合點，

area of the image,

it concludes that there is a face there. This kind of algorithm won't find your

那就會得出那邊有一張臉的結論。這一類的運算法則並不會找到你的臉

face if you're really tilted or facing sideways, but they're really accurate for

如果你的真的很歪或者是看向旁邊的話，但對於臉部正面，他真的很精準

frontal faces, and it's how digital cameras have been putting boxes around

而且電子相機怎麼被放在盒子中已經很久了。

faces for years. But in order to apply this virtual lipstick, the app needs to

但為了申請電子唇膏，比起只是偵查我的臉來說，

do more than just detect my face.

這程式需要做的更多。

It has to locate my facial features.

他必須定位我的臉部特徵，

According to the patents, it does this with an “active shape model” -- a statistical

根據這個專利權，他使用了「動態銳化模式」，一個統計學上關於臉部銳化的模式

model of a face shape that's been trained by people manually marking the

被人類多次手動標記臉部特徵的邊緣，

borders of facial features on hundreds, sometimes thousands of sample images.

有時是上百張，有時是上千張的影像。

The algorithm takes an average face from that trained data and aligns it with the

這個運算法則從那些測試過的資料中取了臉部的平均值，

image from your phone's camera, scaling it and rotating it according to where it

並使用你手機的相機去整理這些圖像，縮放它且旋轉它

already knows your face is located.

根據他對你臉部已知的定位。

But it's not a perfect fit, so the model analyzes the pixel data around each of the points,

但並不是完美的契合，所以這模式會分析每個點附近的像素資料，

looking for edges defined by brightness and darkness. From the training images,

藉由光影來尋找邊緣以給出定義。從這些正被測試的影像中，

the model has a template for what the bottom of your lips should look like,

這模式會有你的下唇看起來應該像什麼的模版，

for example, so it looks for that pattern in your image and adjust the point to match it.

例如，它在你的影像上找尋特定圖案，並且調整點去符合它。

Because some of these individual guesses might be wrong,

因為這之中的個別推測可能會是錯的，

the model can correct and smooth them by taking into account the locations of all

模式可以藉由計算所有其他的落點來修正與修飾他們，

the other points. Once it locates your facial features, those points are used as

一旦它鎖定你的臉部特徵，那些點就會被用作為

coordinates to create a mesh.

座標並創出網格。

That's a 3D mask that can move, rotate, and scale along with your face as the

那就是3D面具可以移動、旋轉以及自行根據你的臉來衡量影像資料進入的原因

video data comes in for every frame and once they've got that, they can do a lot with it.

且一旦他們得到了，他們就可以依此做很多事情。

They can deform the mask to change your face shape, change your eye color,

他們可以把面具變形，以改變你臉部的輪廓、改變你眼睛的顏色、

add accessories, and set animations to trigger when you open your mouth

增加物件，並且設置動畫，當你張開嘴巴，或者是動你的眉毛時，就會觸發。

or move your eyebrows.

And like the IOS app Face Swap Live, Snapchat can switch your face with a

就像蘋果的程式Face Swap Live，Snapchat可以把你的臉換成朋友的

friend's, although that involves a bunch more data.

雖然他牽涉到更大量的資料。

The main components of this technology are not new. What's new is the ability to

這科技的主要組件，其實並沒有很新。新穎的部份是實際在運作的能力，

run them in real time, from a mobile device. That level of processing speed is

在一個移動裝置中。運算速度的程度

a pretty recent development.

是最近才發展起來的。

So why go through all this trouble just to give people a virtual flower crown?

那...為何克服這些所有的問題後，只能讓人們多一頂虛擬的花冠？

Well Snapchats sees a revenue opportunity here. In a world that's

嗯...Snapchat看到一個賺錢的機會。在一個

flooded with advertisements,

充滿廣告的世界裡，

maybe the best hope that brands have to get us to look at their ads... is to

或許品牌想讓我們看到廣告的最好方法就是

put them on our faces.

把廣告放在我們的臉上。

Facial detection has a creepy side too, particularly when it's used to identify you by name.

臉部偵查也是個讓人毛骨悚然的概念，特別是當它被用來以你的名字辨識你的時候。

Both the FBI and private companies like Facebook and Google are massing huge

警察跟私人公司，像是臉書跟Google都集合大量

databases of faces and there's currently no federal law regulating it.

臉不貲要，且他們近期並不受聯邦法律管轄。

So some privacy advocates have come up with ways to camouflage your face from

所有有些隱藏的支持者開始出現，藉由臉部偵查運算法則來偽裝成你的臉。

facial detection algorithms.

It's actually illegal in a lot of places to wear a face mask in public,

在很多地方，在公共場合帶假面具是違法的，

so this project by artist Adam Harvey suggest some things that you can do with

所以這個由Adam Harvey 提議的計畫，有些是你可以幫你的頭髮做的，

your hair and your makeup that can, for now, make your face Invisible to computers.

還有你的妝容，現在呢，讓你的臉無法被電腦辨識出來吧！