Yuze He
Carnegie Mellon University
yuzeh@cs.cmu.edu
Bio
Yuze He is a postdoctoral fellow in Computer Science Department at Carnegie Mellon University, co-advised by Prof. Peter Steenkiste and Prof. Srinivasan Seshan. She earned her Ph.D. from The Chinese University of Hong Kong supervised by Prof. Guoliang Xing, after completing her Bachelor’s degree with honors in Zhejiang University.
Yuze’s research interest spans mobile computing, networking, and systems. Her work primarily explores cooperative autonomous driving, aiming to enhance autonomous vehicle capabilities through smart infrastructure and inter-vehicle collaboration. Her contributions have been recognized with several awards, including the MobiCom 2023 Best Community Contribution Award, the 2024 ACM SIGBED China Chapter Doctoral Dissertation Award, a Gold Prize at the 49th International Exhibition of Inventions in Geneva, the 2023 International Doctoral Forum Best Paper Award, and the N2Women Young Researcher Fellowship.
Areas of Research
- Communications and Networking
- Computer Systems
Towards Cooperative Autonomous Driving: From Perception to HD Mapping
Autonomous driving has the potential to revolutionize urban mobility and road safety in smart cities. However, the mission-critical nature of autonomous driving has resulted in increasingly complex and expensive vehicular systems, which pose significant barriers to widespread adoption in practice. Moreover, many recent accidents have raised public concerns about the reliability and safety of existing autonomous driving systems. An emerging paradigm promising to address this challenge is infrastructure-assisted autonomous driving. This approach utilizes intelligent roadside infrastructure, such as lampposts equipped with sensors and computing units, to provide real-time services and information to autonomous vehicles.
My research focuses on advancing infrastructure-assisted autonomous driving, with a vision of enhancing autonomous driving through the sensing and computing capabilities of roadside smart infrastructure. I have developed several vehicle-infrastructure cooperative systems targeting key aspects of autonomous driving: perception, localization, and high-definition (HD) mapping. Specifically, we capitalize on the broader and unobstructed field of view offered by infrastructure to augment vehicle perception. We also utilize the stationary characteristics of infrastructure to assist in vehicle localization. Furthermore, we propose leveraging the advantages of stationary roadside infrastructure, such as cumulative observations, to construct accurate, up-to-date, and scalable HD maps for autonomous driving. My research has established successful collaborations with over six groups across academia and industry.