Overview
Deep neural networks (DNNs) are increasingly used in time-critical, learning-enabled cyber-physical applications such as autonomous driving and robotics. Many of them have stringent temporal (i.e., “real-time”) requirements. Despite the growing use of various deep learning models in real-time systems, protecting DNN inference from adversarial threats while preserving model privacy and confidentiality remains a key concern for resource and timing-constrained systems. Based on our recent exploration and findings in this domain, this talk will present challenges and new techniques to enable confidential deep neural for real-time systems.
Speaker Bio
Dr. Monowar Hasan is a Computer Science Assistant Professor at Washington State University (WSU). Before joining WSU, he held an Assistant Professor position at Wichita State University from 2021-2022. Dr. Hasan received his Ph.D. in Computer Science from the University of Illinois at Urbana-Champaign (UIUC) in 2020. His research interests include exploring security and resiliency techniques of cyber-physical system domains.