University of Washington researchers have developed new algorithms that can turn audio clips into a realistic, lip-synced video of the person speaking those words.
As detailed in a paper to be presented August 2 at SIGGRAPH 2017 in L.A., the team successfully generated realistic video of former president Barack Obama talking about terrorism, fatherhood, job creation and other topics using audio clips of those speeches and existing weekly video addresses that were originally on a different topic.
Ira Kemelmacher-Shlizerman, an assistant professor at the UW's Paul G. Allen School of Computer Science & Engineering said, Realistic audio-to-video conversion has practical applications like improving video conferencing for meetings, as well as futuristic ones such as being able to hold a conversation with a historical figure in virtual reality by creating visuals just from audio.
In a visual form of lip-syncing, the system converts audio files of an individual's speech into realistic mouth shapes, which are then grafted onto and blended with the head of that person from another existing video.
In the future video, chat tools like Skype or Messenger will enable anyone to collect videos that could be used to train computer models, Kemelmacher-Shlizerman said.
Because streaming audio over the internet takes up far less bandwidth than video, the new system has the potential to end video chats that are constantly timing out from poor connections.
When you watch Skype or Google Hangouts, often the connection is stuttery and low-resolution and really unpleasant, but often the audio is pretty good, said co-author and Allen School professor Steve Seitz. So if you could use the audio to produce much higher-quality video, that would be terrific.
By reversing the process feeding video into the network instead of just audio the team could also potentially develop algorithms that could detect whether a video is real or manufactured.
The new machine learning tool makes significant progress in overcoming what's known as the uncanny valley problem, which has dogged efforts to create realistic video from audio. When synthesised human likenesses appear to be almost real but still manage to somehow miss the mark people find them creepy or off-putting.
People are particularly sensitive to any areas of your mouth that don't look realistic, said lead author Supasorn Suwajanakorn, a recent doctoral graduate in the Allen School. If you don't render teeth right or the chin moves at the wrong time, people can spot it right away and it's going to look fake. So you have to render the mouth region perfectly to get beyond the uncanny valley.
A neural network first converts the sounds from an audio file into basic mouth shapes. Then the system grafts and blends those mouth shapes onto an existing target video and adjusts the timing to create a new realistic, lip-synced video.
Previously, audio-to-video conversion processes have involved filming multiple people in a studio saying the same sentences over and over to try to capture how a particular sound correlates to different mouth shapes, which is expensive, tedious and time-consuming. By contrast, Suwajanakorn developed algorithms that can learn from videos that exist in the wild on the internet or elsewhere.
There are millions of hours of video that already exist from interviews, video chats, movies, television programs and other sources. And these deep learning algorithms are very data hungry, so it's a good match to do it this way, Suwajanakorn said.
Rather than synthesising the final video directly from audio, the team tackled the problem in two steps. The first involved training a neural network to watch videos of an individual and translate different audio sounds into basic mouth shapes.
By combining previous research from the UW Graphics and Image Laboratory team with a new mouth synthesis technique, they were then able to realistically superimpose and blend those mouth shapes and textures on an existing reference video of that person. Another key insight was to allow a small time shift to enable the neural network to anticipate what the speaker is going to say next.
The new lip-syncing process enabled the researchers to create realistic videos of Obama speaking in the White House, using words he spoke on a television talk show or during an interview decades ago.
Currently, the neural network is designed to learn on one individual at a time, meaning that Obama's voice speaking words he actually uttered is the only information used to drive the synthesised video. Future steps, however, include helping the algorithms generalise across situations to recognise a person's voice and speech patterns with less data with only an hour of video to learn from, for instance, instead of 14 hours.
The research was funded by Samsung, Google, Facebook, Intel and the UW Animation Research Labs.
A neural network first converts the sounds from an audio file into basic mouth shapes. Then the system grafts and blends those mouth shapes onto an existing target video and adjusts the timing to create a new realistic, lip-synced video.
Most recent headlines
04/08/2024
Dalet, a leading technology and service provider for media-rich organizations, is excited to announce Santiago Solanas as its new Chief Executive Officer (CEO)....
03/06/2024
Dalet, a leading technology and service provider for media-rich organizations, a...
04/05/2024
Spotify Wrapped Campaign Spot Graded With DaVinci Resolve Studio
Brie Clayton May 3, 2024
0 Comments
Blackmagic Design today announced that Colorist M...
04/05/2024
Three Boston Conservatory at Berklee Alums Nominated for Tony Awards An additional eight alums and two current students performed in nominated productions.
...
03/05/2024
An innovative learning tool to help media and civil society better understand ho...
03/05/2024
PARK CITY, UTAH - JANUARY 21: (L-R) Actors K stutis Cic nas and Greta Grinevi i t , Director Marija Kavtaradze and Producer Marija Razgute attend the 2023 Sunda...
03/05/2024
If you aren't familiar with Danny Ocean's music, it's only a matter ...
03/05/2024
Spotify has an unwavering commitment to supporting emerging artists across all genres, to helping them launch and thrive in their careers, and to connecting the...
03/05/2024
Como uno de los sonidos de mayor expansi n en el mundo, la M sica Mexicana va m ...
03/05/2024
As one of the fastest-growing sounds worldwide, M sica Mexicana is more than jus...
03/05/2024
NEW YORK FuboTV reported that it lost 110,000 subs in Q1 2024 but it exceeded its guidance for the quarter with double digit year-over-year increases across key...
03/05/2024
Only 3 Weeks Left To Enter The Creative Industries Most Award
Brie Clayton May 3, 2024
0 Comments
Creativepool is giving a last call to the creative i...
03/05/2024
WePlay Studios Transforms the Future of Live Event Storytelling with AJA Gear
Brie Clayton May 3, 2024
0 Comments
By 2032, the esports market is expec...
03/05/2024
WASHINGTON Shannon Bream, Fox News Sunday anchor and chief legal correspondent, sat for a keynote chat at The Business of TV News, with Tom Umstead, senior co...
03/05/2024
WASHINGTON Asked how important is local news, Martha Raddatz, who has covered foreign conflicts for decades and moderated presidential debates for ABC News, s...
03/05/2024
Sony Pictures and Apollo Global Management have made a $26 billion cash offer fo...
03/05/2024
CBS announced its 2024-2025 primetime schedule, which features the new dramas NCIS: Origins, Matlock and Watson, and new comedies Poppa's House and Georgie ...
03/05/2024
Sports-focused streaming service Fubo said it reduced its red ink in the first quarter, despite losing 110,000 subscribers since the end of the year....
03/05/2024
Walmart is reportedly about to release a new 4K streaming box/smart speaker hybrid device for Google TV that could retail for as low as $50, according to a repo...
03/05/2024
Enhancements to its Delivery Workspace solution provide increased content delivery robustness and security
By Matthew Corrigan
Published: May 3, 2024
Enha...
03/05/2024
According to reports, if the bid is successful, Sony would be the significant majority shareholder of the new private company
By Jenny Priestley
Published: M...
03/05/2024
Located in Manchesters Spinningfields district, the course aims to equip students with the skills, knowledge and industry insight for a career in the creative s...
03/05/2024
Lightbridge has created a new Cine Reflect Lighting System kit, dedicated to enabling artistic lighting with consideration of today's productivity practical...
03/05/2024
Broadcast media systems integrator BeckTV and Lawo worked together on a large-scale AES67 build of radio control rooms at KPBS, San Diego's NPR and PBS memb...
03/05/2024
Raleigh, NC, April 12, 2024 SmallHD announces PageOS 6, a powerful software update free for owners of SmallHD monitors, featuring increased functionality, exp...
03/05/2024
SmallHD announces the Quantum 32 a new 31.5 OLED monitor designed to provide reference HDR/SDR images for post-production final color grading, as well as cri...
03/05/2024
By 2032, the esports market is expected to grow to $9.29 billion, bolstered by a global player count and fan following that both continue to climb. The 2023 Lea...
03/05/2024
Amagi, the global leader in cloud-based SaaS technology for broadcast and connected TV (CTV), today announced that the 11th edition of the Amagi Global FAST Rep...
03/05/2024
Riedel Communications today announced that Westdeutscher Rundfunk (WDR) is relying on a media and intercom network based on Riedel technology for the renewal of...
03/05/2024
Broadpeak , a leading provider of content delivery network (CDN) and video streaming solutions for content providers and pay-TV operators worldwide, announced t...
03/05/2024
Media technology innovator Limecraft will exhibit the latest updates to its online platform on stand G79 at the 15-16 May 2024 Media Production & Technology Sho...
03/05/2024
Audio interface specialist Prism Sound is partnering with console manufacturer Solid State Logic, professional monitor manufacturer PMC and acoustic panel and d...
03/05/2024
Intinor, Sweden's leading developer of products and solutions for high-quality video over the internet, is set to join forces with its UK partner Zest Techn...
03/05/2024
NAKIVO Inc., a fast-growing software company dedicated to protecting physical, virtual, cloud, and SaaS environments, announced its Q1 2024 financial results to...
03/05/2024
Hiltron Communications announces an addition to its range of satellite communication products and systems: a field-upgradable motorisation kit specifically desi...
03/05/2024
Visit LiveU at Stand D30
With this year an exceptional one for major sporting events alongside multiple, key elections, the power and flexibility of LiveU'...
03/05/2024
Agile Content, the leader in the development of technology and solutions for the provision of television services over the Internet, announces Koldo Unanue as i...
03/05/2024
Eight co-signers representing streaming services, pay TV operators and public interest groups have sent a letter to Congressional leaders urging their Committee...
03/05/2024
BOSTON Brightcove has announced that it has implemented Amazon Q Business, a new generative AI assistant on Amazon Web Services (AWS)....
03/05/2024
NEW YORK CBS has declared that it will finish the 2023-24 season as Americas Most-Watched Network in primetime for the 16th consecutive season....
03/05/2024
A new survey indicates that U.S. businesses and consumers had an overall positive attitude towards their technology spending plans in the first quarter of 2024,...
03/05/2024
Sony and Apollo Global Management have sent an offer letter to Paramount Global's board of directors making an $26 billion all cash offer for Paramount Glob...
03/05/2024
EMG / Gravity Media has appointed Charlie Cubbon chief operating officer (COO) and Jamie Hindhaugh regional CEO for the U.K., U.S., Australia and the Middle Eas...
03/05/2024
Final 2024 Season Games to Air Saturdays on WRAL at 10:30am in May, Championship on June 1
Taping for the final episodes of the 2024 season of WRAL-TV's Br...
03/05/2024
Estrella MediaCo said it is expanding its connected TV portfolio by launching three new channels with Curiosity that will first appear on Samsung TV Plus....
03/05/2024
Nielsen said that actors' union SAG-AFTRA has selected Nielsen as its third-party provider of streaming content measurement....
03/05/2024
Cineverse said it has released a public beta version of cineSearch, its artificial intelligence-based system designed to help users find content to watch....
03/05/2024
Court TV said that its series Accomplice to Murder with Vinnie Politan is returning for a second season beginning May 5....
03/05/2024
Paramount Advertising said it is working with Mastercard and research and analytics company EDO to measure the impact of advertising campaigns and improve their...
University of Washington researchers have developed new algorithms that can turn audio clips into a realistic, lip-synced video of the person speaking those words. 
