Supasorn Suwajanakorn: Fake videos of real people -- and how to spot them
蘇帕索恩蘇瓦耶納柯恩: 真人的假影片以及如何辨視出它們
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.
歐巴馬是真的?
refinance their homes,
協助家庭做房款再融資,
like high-tech manufacturing,
that creates good new jobs.
我們如何走到這一步。
chance for learning about the Holocaust
保存我們從大屠殺生還者那裡
interactive conversations
of a real Holocaust survivor.
survive the Holocaust?
從大屠殺活下來?
were prerecorded in a studio.
在攝影棚預先錄影好的。
and to him as a person.
及他本人很有連結。
about human interaction
or movies could ever teach us.
沒有辦法教導我們的。
like this for anyone?
創造這樣的模型?
and acts just like them?
說話方式,和行為的模型?
using nothing but these:
就能建造出一個人的模型:
this kind of passive information,
到所有人的關鍵。
a Nobel Prize winner in physics
if we could bring him back
and inspire millions of kids,
but in any language?
還可以用任何語言?
for advice and hear those comforting words
還能問問他們意見,
book authors, alive or not,
讓不論還活著或已故的作家
for anyone interested.
給想要聽的讀者聽。
here are endless,
3D face model from any image
一個人非常細節的 3D 面部模型,
from different views.
on each video frame
output model from different angles.
is very challenging,
is that we are going to analyze
of the person beforehand.
we can just search on Google,
我們用 Google 搜尋就可以了,
to build an average model,
一個平均的模型,
to recover the expression
它能夠很精密地重現出表情,
like creases and wrinkles.
can come from your typical photos.
就可以做到了。
what expression you're making
that there are a lot of them.
a new blending technique
一種新的混和技術,
a single averaging method
facial textures and colors.
面部質感和顏色。
of a model of a person,
is by a sequence of static photos.
一系列靜態照片來進行。
depending on the expression.
皺紋有時會出現有時會消失。
to drive the model.
對,但不知何故
some more amazing people.
are controllable models
from their internet photos.
the motion from the input video,
It's a difficult bill to pass,
這個法案很困難,
is to capture their mannerisms
其實是要捕捉他們的動作習性,
of these people talks and smiles.
和微笑的獨特方式。
actually teach the computer
video footage of the person?
I let a computer watch
giving addresses.
內容全是歐巴馬在演講。
given only his audio.
我們就能產生出這樣的成果。
14.5 million new jobs
美國的企業已經創造出了
is only the mouth region,
使用的是類神經網路,
into these mouth points.
or through Medicare or Medicaid.
或醫療保險、醫療補助來達成。
enhance details and teeth,
強化細節和牙齒,
and background from a source video.
以及來源影片的背景上。
女性能接受免費檢查,
just for being a woman.
就要付比較多錢。
on a parent's plan until they turn 26.
掛在一位家長的方案底下。
seem very realistic and intriguing,
非常真實且有意思,
frightening, even to me.
of a person, not to misrepresent them.
人類模型,而不是要故意歪曲他。
is its potential for misuse.
是這項工具可能被誤用。
about this problem for a long time,
在市場上之後,
first hit the market.
在想著這個問題。
on countermeasure technology,
我也在研究反制的技術,
effort at AI Foundation,
基金會的一項計畫,
of machine learning and human moderators
is called Reality Defender,
叫做「真相守衛者」,
that can flag potentially fake content
能標出有可能是仿冒的內容,
驗證之前,傷害就已造成,
that we make everyone aware
要讓大家都能意識到,
and be critical about what we see.
並對所看到的內容抱持批評的態度。
we can fully model individual people
還有很長的一段路要走,
the safety of this technology.
這項科技的安全性。
小心地使用它,
positive impact on the world
the way we want it to be.
來型塑我們的未來。
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