Carnegie Mellon University

Hamerschlag Hall

April 01, 2019

Savvides on AI and facial recognition

Marios Savvides spoke with CNET in an article about how AI has helped to drastically improve facial recognition. While there are privacy and bias concerns, facial recognition is now being used more widely, at airports, in home security systems, and on cruises, with a 99.7 percent accuracy for the most cutting-edge systems. However, even deep learning neural networks can make mistakes. Savvides, director of the CyLab Biometrics Center, separates some of the data to make things clearer for the neural net. His team can reconstruct faces even in conditions that aren’t optimal. “We live in a time where AI can surpass the human brain’s capability,” he says. Read the article here.