Email Login
Community
News | Research | DeepLearning | Seminar | Q&A

AI Creates Fake Obama...!

GreatMind 2017.07.14 13:05 조회 수 : 261

AI Creates Fake Obama

By Charles Q. Choi

Posted 12 Jul 2017 | 14:00 GMT

 

Two pictures of Barack Obama side-by-sidePhoto: University of Washington.

Artificial intelligence software could generate highly realistic fake videos of former president Barack Obama using existing audio and video clips of him, a new study [PDF] finds.

 

Such work could one day help generate digital models of a person for virtual reality or augmented reality applications, researchers say.

 

Computer scientists at the University of Washington previously revealed they could generate digital doppelgängers of anyone by analyzing images of them collected from the Internet, from celebrities such as Tom Hanks and Arnold Schwarzenegger to public figures such as George W. Bush and Barack Obama. Such work suggested it could one day be relatively easy to create such models of anybody, when there are untold numbers of digital photos of everyone on the Internet.

 

The researchers chose Obama for their latest work because there were hours of high-definition video of him available online in the public domain. The research team had a neural net analyze millions of frames of video to determine how elements of Obama's face moved as he talked, such as his lips and teeth and wrinkles around his mouth and chin.

 

 

In an artificial neural network, components known as artificial neurons are fed data, and work together to solve a problem such as identifying faces or recognizing speech. The neural net can then alter the pattern of connections among those neurons to change the way they interact, and the network tries solving the problem again. Over time, the neural net learns which patterns are best at computing solutions, an AI strategy that mimics the human brain.

 

In the new study, the neural net learned what mouth shapes were linked to various sounds. The researchers took audio clips and dubbed them over the original sound files of a video. They next took mouth shapes that matched the new audio clips and grafted and blended them onto the video. Essentially, the researchers synthesized videos where Obama lip-synched words he said up to decades beforehand.

 

The researchers note that similar previous research involved filming people saying sentences over and over again to map what mouth shapes were linked to various sounds, which is expensive, tedious and time-consuming. In contrast, this new work can learn from millions of hours of video that already exist on the Internet or elsewhere.

 

 

One potential application for this new technology is improving videoconferencing, says study co-author Ira Kemelmacher-Shlizerman at the University of Washington. Although teleconferencing video feeds may stutter, freeze or suffer from low-resolution, the audio feeds often work, so in the future, videoconferencing may simply transmit audio from people and use this software to reconstruct what they might have looked like while they talked. This work could also help people talk with digital copies of a person in virtual reality or augmented reality applications, Kemelmacher-Shlizerman says.

 

The researchers note their videos are currently not always perfect. For example, when Obama tilted his face away from the camera in a target video, imperfect 3-D modeling of his face could cause parts of his mouth to get superimposed outside the face and onto the background.

 

In addition, the research team notes their work did not model emotions, and so Obama's facial expressions in the output videos could appear too serious for casual speeches or too happy for serious speeches. However, they suggest that it would be interesting to see if their neural network could learn to predict emotional states from audio to produce corresponding visuals.

The researchers were careful to not generate videos where they put words in Obama's mouth that he did not at some other time utter himself. However, such fake videos are “likely possible soon,” says study lead author Supasorn Suwajanakorn, a computer scientist at the University of Washington. 

However, this new research also suggests ways to detect fake videos in the future. For instance, the video manipulation the researchers practiced can blur mouths and teeth. “This may be not noticeable by human eyes, but a program that compares the blurriness of the mouth region to the rest of the video can easily be developed and will work quite reliably,” Suwajanakorn says.

 

The researchers speculated that the link between mouth shapes and utterances may be to some extent universal for people. This suggests that a neural network trained on Obama and other public figures could be adapted to work for many different people.

 

The research was funded by Samsung, Google, Facebook Intel and the University of Washington. The scientists will detail their findings [PDF] on Aug. 2 at the SIGGRAPH conference in Los Angeles.

 

Link: http://spectrum.ieee.org/tech-talk/robotics/artificial-intelligence/ai-creates-fake-obama?utm_campaign=TechAlert_07-13-17&utm_source=Tech+Alert&bt_alias=eyJ1c2VySWQiOiAiMjNhZTU1YzktY2Y2Yy00YjM4LThkNzctYWVlMjVkOWY1ZGFjIn0%3D&utm_medium=Email

번호 제목 글쓴이 날짜 조회 수
공지 2019 IVPL 송년회 [1] yjchoi 2019.12.20 265
공지 [한국과총]글로벌 현장연수 참가자 모집 GreatMind 2018.04.04 292
공지 Inform Your Students: Global Student Challenge. Over US$2,000 in Prizes! GreatMind 2017.01.12 4343
공지 nVidia Parallel Computing Software Engineer 채용 GreatMind 2016.11.16 717
87 Google Is Helping the Pentagon Build AI for Drones GreatMind 2018.03.09 539
86 카이스트, 모바일용 딥 러닝 AI반도체 개발 GreatMind 2018.03.06 690
85 [News] Congratulation...!!! Dr. Kalyan's paper has been accepted for publication in the IEEE Transactions on Industrial Electronics...!!! GreatMind 2018.03.02 1180
84 [News] Mr. Kim Hyun-Teak visited to our lab. to collaborate with us..!!! [8] file GreatMind 2018.02.12 644
83 [News] 성경식품 UCC 아이디어 공모전 동상 수상....! GreatMind 2018.01.11 85
82 Building AI systems that work is still hard..! [3] GreatMind 2018.01.04 1527
81 [Notice] Dr. Kalyan's paper has been accepted as minor revision in IEEE Transactions on Industrial Electronics...!!! GreatMind 2017.12.26 759
80 JPL's AI-Powered Racing Drone Challenges Pro Human Pilot GreatMind 2017.12.20 733
79 [Notice] 캡스톤디자인팀 똑띠 학부우수 논문상 수상 (DdokDdi has won the Outstanding Paper Award in KMMS Autumn conference)...!!!! file GreatMind 2017.11.29 204
78 [Notice] Young-Ju Choi, Ji-Hae Kim's paper has been accepted for publication in Journal of Digital Contents Society..!!!! GreatMind 2017.11.21 65
77 Carnegie Mellon Solves 12-Year-Old DARPA Grand Challenge Mystery..!!! GreatMind 2017.10.26 1054
76 NVIDIA Jetson Developer Contest...!!! GreatMind 2017.10.24 61
75 [Notice] Mr. Young-Woon Lee and Ms. Young-Ju Choi's paper has been accepted for publication in Journal of Digital Contents Society..!!!! GreatMind 2017.10.10 54
74 [Notice] Mr. Lee Jong-Hyeok of our group, his biography has been listed (accepted) on Who's Who in the World 2018....!!! GreatMind 2017.09.08 327
73 [Notice] Mr. Lee Jong-Hyeok has been nominated in Marquis Who's Who in the World ...!! GreatMind 2017.09.07 1055
72 [Notice] Mr. Young-Woon Lee's paper has been accepted for presentation in IEEE ICCE 2018..!!! GreatMind 2017.09.06 22
71 2017년 실리콘밸리 인턴십 운영지원사업 참여학생 선발 공고 GreatMind 2017.09.05 281
70 Smart glasses let you turn off the lights in the blink of an eye GreatMind 2017.08.01 339
69 Facebook, Microsoft, and IBM Leaders on Challenges for AI and Their AI Partnership..! GreatMind 2017.07.14 105
» AI Creates Fake Obama...! GreatMind 2017.07.14 261