University of Washington researchers have developed new algorithms that solve a thorny challenge in the field of computer vision: turning audio clips into a realistic, lip-synced video of the person speaking those words.
As detailed in a paper to be presented Aug. 2 at SIGGRAPH 2017, the team successfully generated highly-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.
“These type of results have never been shown before,” said Ira Kemelmacher-Shlizerman, an assistant professor at the UW’s Paul G. Allen School of Computer Science & Engineering. “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. This is the kind of breakthrough that will help enable those next steps.”
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.
The team chose Obama because the machine learning technique needs available video of the person to learn from, and there were hours of presidential videos in the public domain. “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 synthesized 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.”
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 synthesizing 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 synthesized video. Future steps, however, include helping the algorithms generalize across situations to recognize 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.
“You can’t just take anyone’s voice and turn it into an Obama video,” Seitz said. “We very consciously decided against going down the path of putting other people’s words into someone’s mouth. We’re simply taking real words that someone spoke and turning them into realistic video of that individual.”
Learn more: Lip-syncing Obama: New tools turn audio clips into realistic video
The Latest on: Audio-to-video conversion
[google_news title=”” keyword=”computer vision” num_posts=”10″ blurb_length=”0″ show_thumb=”left”]- What Is Machine Vision and How Important Is It to Self-Driving Cars?on May 17, 2024 at 5:32 am
Machine vision is a technology that allows machines to see and recognize objects in their surrounding environment. It is a subset of computer vision that focuses on the industrial applications of ...
- Google I/O 2024: DeepMind Showcases Real-Time Computer Vision-Based AI Interaction With Project Astraon May 15, 2024 at 6:59 am
Google’s Project Astra showed an AI chatbot using a smartphone camera to describe the environment in real time.
- Global AI in Computer Vision Market Size To Exceed USD 118.56 Billion By 2033| CAGR of 21.48%on May 15, 2024 at 2:00 am
The Global AI in Computer Vision Market Size was Valued at USD 16.94 Billion in 2023 and the Worldwide AI in Computer Vision Market Size is Expected to Reach USD 118.56 Billion by 2033, according to a ...
- Google Teases Computer Vision, Conversational Capabilities of Gemini AI Ahead of Google I/O Eventon May 14, 2024 at 6:20 am
Google’s Gemini also appears to be better at holding a conversation Google I/O event is scheduled for 10:30pm IST on May 14 Google can announce new AI features for Gemini during the event ...
- The Apple Vision Pro may have tanked — but spatial computing is still the future, Deloitte sayson May 9, 2024 at 10:42 pm
Spatial computing, or the blending of physical and digital worlds, is not just a sci fi pipe dream — it's the next level of interaction.
- 6 Ways Computer Vision is Re-envisioning the Future of Drivingon May 9, 2024 at 8:06 am
Today’s cars are like supercomputers on wheels – smarter, safer, faster, and more personalized thanks to technological advances. One transformative innovation steering this revolution is computer ...
- Is Your Computer Giving You a Headache? The Alarming Rise of Computer Vision Syndromeon May 7, 2024 at 11:48 pm
Struggling with eye strain or headachesafter screen time? Learn the causes, symptoms & simple solutions to protect your eyes in digital age.
- Increasing Prevalence of Computer Vision Syndrome Highlights the Need for Preventive Vision Careon May 6, 2024 at 2:27 am
Although rapid technological development in recent years has helped enhance daily productivity and efficiency it has significantly contributed to adverse consequences for the population eye health Spe ...
- How Computer Vision Is Transforming Cybersecurityon April 27, 2024 at 1:40 am
I write about contemporary cybersecurity and online privacy issues. Integrating computer vision technology is a big step forward in cybersecurity defense strategies. With its ability to automate ...
- Computer vision researcher develops privacy software for surveillance videoson April 24, 2024 at 5:00 pm
Computer vision researcher develops privacy software for surveillance videos. ScienceDaily . Retrieved May 15, 2024 from www.sciencedaily.com / releases / 2024 / 04 / 240425131606.htm ...
via Google News and Bing News