Smith receives Google grant to investigate privacy issues

Rebekka Coakley
September 24, 2015

UNIVERSITY PARK, Pa. – Adam Smith, associate professor of computer science and engineering in Penn State's  School of Electrical Engineering and Computer Science, has received a grant from Google to investigate privacy issues in a state-of-the-art training technique called deep learning.

The grant, which is shared with Vitaly Shmatikov at Cornell, will support investigations into what deep learning systems can leak about sensitive inputs, as well as the development of a system for privacy-preserving deep learning.

Google Faculty Research Awards support cutting-edge research in computer science, computer engineering and related fields. Approximately 200 grants are awarded globally each year. Smith and Shmatikov will share their results with relevant research groups at Google and have opportunities to collaborate with those groups, but all the products of the research will be publicly available.

“Deep learning is already widely used, especially at Google, to recognize speech and images and to do lots of other things like drive cars,” said Smith. “Deep learning is often applied to very sensitive data, however. We are interested in making sure that sensitive, private data remain private.”

Smith explained that deep learning based on artificial neural networks has led to dramatic improvements in speech and image recognition, language translation and other artificial intelligence tasks, but added that the collection of training data from millions of users presents serious privacy risks. He is working on implementing a system that will enable many data holders to collaborate to learn accurate neural-network models without sharing their training datasets or leaking sensitive information about them.

According to Google, Faculty Research Awards are given to projects that might especially benefit from collaboration with Google. Here is a complete list of this year’s recipients.

(Media Contacts)

Last Updated September 25, 2015