Published January 3, 2023
Siwei Lyu, a professor in the University at Buffalo Department of Computer Science and Engineering (CSE) and one of the world’s leading experts on deepfakes, has been named a distinguished member of the Association for Computing Machinery (ACM).
The world’s largest computing society, ACM recognizes up to 10% of its worldwide membership as distinguished members for significant impacts on the field of computing. This year’s 67 inductees work at leading universities, corporations and research institutions in 16 different countries, and have made advancements in areas like cybersecurity, software engineering and computer science education.
“I am honored and humbled to be chosen as one of this year's ACM distinguished members,” Lyu says. “It is an exciting recognition of my research contributions and career up to this point.”
Lyu, SUNY Empire Innovation Professor, has made significant contributions to combating misinformation and disinformation online. His research focuses on media forensics, detecting and mitigating manipulated or fabricated digital media like photos and videos.
Lyu is director of UB’s Media Forensics Lab and founding co-director of UB’s Center for Information Integrity, the latter of which was launched this year to accelerate the university’s multidisciplinary research on disinformation. Lyu and other researchers affiliated with the center are leading a $5 million National Science Foundation (NSF) project to create digital tools that help older adults better recognize and protect themselves from online scams and other disinformation.
As a professor at the University at Albany in 2018, Lyu developed the first algorithm for detecting synthetic videos of human faces made with artificial intelligence, known as deepfakes. His method determines whether videos are fake based on a lack of realistic blinking.
Lyu has become a go-to expert on deepfakes and other manipulated media. He has authored several of the most-cited works in this area, testified before the U.S. House of Representatives, and been quoted extensively by prominent media outlets like CNN, BBC and The Washington Post.
USA Today recently consulted Lyu on fact-checking stories about a doctored photo of an upside-down city propagating the erroneous flat-Earth theory, as well as a digital rendering of Mars intended to cast doubt on whether exploration missions to the planet really happened.
Publications, including The Independent, have also turned to Lyu on the rise of misinformation targeting politicians. He has offered his expertise on a doctored Facebook photo of a Costco screen blaming President Joe Biden for high gas prices, as well as a manipulated TikTok video of former President Donald Trump appearing to defend rapper Kanye West.
“Since joining us in 2020, Siwei has enhanced our department’s national reputation as a source for detecting and mitigating disinformation and misinformation online,” says Jinhui Xu, professor and chair of CSE. “His work is at the cutting edge of a field that has large ramifications for ensuring that everyone has equal access to accurate information.”
The ACM distinguished member honor caps off a year of recognition for Lyu. He was named a fellow of both the International Association for Pattern Recognition and the Institute of Electrical and Electronic Engineers (IEEE) earlier this year. His previous honors include an IEEE Region 1 Technological Innovation (Academic) Award, Google Faculty Research Award, SUNY Chancellor’s Award for Excellence in Research, and NSF CAREER Award.
Lyu is currently researching how to expose deepfakes that are generated in real time and could be used to impersonate participants in Zoom meetings.
“The proliferation of misinformation in our information ecosystem erodes our trust in online information,” Lyu says. “It will take a combined effort from academics, researchers, government agencies, social media companies, public media and every online user to address this complex problem.”
UB is actively monitoring developments related to COVID-19. Please check with event organizers for the most up-to-date information about scheduled events.