The colorful frames that dodge recognition. (Carnegie Mellon University)
Carnegie Mellon University
Special Frames Thwart Recognition Software
Carnegie Mellon University College of Engineering and School of Computer Science researchers have created eyeglass frames that can fool commercial-grade facial recognition software. The software will identify “the wrong person with up to 100 percent success rates.”
[STEMRules news that feature Carnegie Mellon University.]
The researchers used the eyeglass frame to fool Alibaba, the Chinese e-commerce giant’s “smile-to-pay” feature which was launched with fanfare in 2015.
[Watch Why I Joined CMU.]
The research paper, Accessorize to a Crime: Real and Stealthy Attacks on State-of-the-Art Face Recognition, of CMU’s was presented on Oct. 28 at the 23rd ACM Conference on Computer and Communications Security, in the Hofburg Palace, Vienna, Austria. ACM is the Association for Computing Machinery.
Meet a co-author of the facial recognition hack paper.
Sruti Bhagavatula
Bhagavatula is a PhD student in the School of Computer Science and CyLab at Carnegie Mellon University. Cyblab, founded in 2003, is one of the largest university-based cyber security research and education centers
Her research interests in security and privacy includes machine learning, the Internet of Things, networks and mobile security. Bhagavatula received her BS in Computer Science from the University of Illinois at Chicago.
She has interned at the MIT Lincoln Laboratory in the Cyber Systems Assessments Group. She also worked in Amazon as a Software Development Engineer (SDE).
An SDE applies the principles and techniques of computer engineering, information analysis, and computer science to design, build, and develop their employers’ computer systems and software.
