N. Asokan, Buse Gul Atli, Sebastian Szyller (University of Waterloo and Aalto University)

"Extraction of Complex DNN Models: Real Threat or Boogeyman?"

N. Asokan

Buse Gul Atli

Sebastian Szyller

Zum Zoom Webinar (Code: 022671)



The success of deep learning in many application domains has been nothing short of dramatic. The success has brought the spotlight onto security and privacy concerns with deep learning. One of them is the threat of "model extraction": when a machine learning model is made available to customers via an inference interface, a malicious customer can use repeated queries to this interface and use the information gained to construct a surrogate model. In this talk, I will describe our work in exploring whether model extraction constitutes a realistic threat. I will also discuss possible countermeasures and the challenges in deploying them in popular machine learning configurations like federated learning.



N. Asokan is a professor of computer science and a David R. Cheriton Chair at the University of Waterloo. He is also an adjunct professor of computer science at Aalto University. His research interests are broadly in the domain of systems security with particular emphases on platform security, and the interplay between artificial intelligence and security/privacy problems. Asokan joined academia after a long career in industrial research, first at IBM, and subsequently at Nokia. He is a fellow of both IEEE and ACM. For more information about his research, visit his homepage or follow him on Twitter at @nasokan.

Buse Gul Atli is a senior doctoral student in the Secure Systems Group. She obtained her M.Sc. degree in Signal, Speech and Language Processing from Aalto University in 2017. She was an intern in Nokia Bell Labs and worked on designing machine learning methods for cybersecurity. She is mainly working on both offensive and defensive methods related to the security and privacy of machine learning.

Sebastian Szyller is a senior doctoral student in the Secure Systems Group. Sebastian is an expert on the security and privacy of AI. He holds a M.Sc. degree in Machine Learning and Data Mining from Aalto University. Prior to joining Aalto, Sebastian worked as a software engineer in investment banking where he designed and implemented high throughput systems that facilitate trading.

User login

Enter your username and password here in order to log in on the website