Huili Chen

Huili Chen is an Assistant Professor in the Department of Electrical and Computer Engineering at the University of California San Diego's Jacobs School of Engineering. Her research spans the intersection of hardware security, deep learning, and human-robot interaction, with a particular focus on intellectual property protection for AI systems and socially interactive robotics for education and family settings.

Her research interests focus on hardware-software co-design for secure and robust deep learning systems, with specific expertise in neural network watermarking, hardware Trojan detection, and privacy-preserving AI. She has made significant contributions to federated learning security, developing frameworks like GALU for logic unlocking and AdaTest for hardware Trojan detection. In human-robot interaction, she investigates long-term multi-person interactions, particularly in home environments with children and parents, exploring how robots can enhance engagement and learning through adaptive role-playing.

Her publication trends reveal a dual research trajectory: one branch focused on deep learning security and hardware co-design (accounting for approximately 60% of her recent work), and another dedicated to socially assistive robotics and human-robot interaction (40%). The security research often involves innovative approaches combining reinforcement learning with hardware constraints, while her HRI work emphasizes longitudinal studies of robot-child-parent dynamics in naturalistic settings.

Dr. Chen actively collaborates with leading researchers including Farinaz Koushanfar at UC San Diego and Cynthia Breazeal at MIT Media Lab. Her work has been published in top-tier venues including IEEE Transactions on Affective Computing, ACM Transactions on Embedded Computing Systems, International Conference on Computer Vision (ICCV), and the ACM/IEEE International Conference on Human-Robot Interaction (HRI).

She is a key contributor to the DAMI-P2C project, which has developed datasets and models for analyzing parent-child multimodal interactions, and leads research on hardware security frameworks for deep neural networks. Her laboratory combines expertise in computer architecture, machine learning, and social robotics to develop systems that are both technically secure and socially effective.