High Performance Computing and Image Processing under Complex Conditions

The Autonomous Driving Research Center conducts research on many key scientific problems including 1. Modeling of learning-based perception systems, 2. Formal modeling of decision-making and planning systems, 3.Trustworthiness design and validation of autonomous driving systems, 4, Trustworthy testing and evaluation system design for autonomous systems. The technical developement if the center has been focusing on the trustworthiness mechanisms of autonomous driving systems, the design and analysis of trustworthy autonomous driving systems, and high-trustworthy datasets and simulation evaluation platforms. The center has achieved breakthroughs in common key technologies such as the trustworthy design and verification technoliogies of autonomous driving systems, and the integration of simulation and datasets for reliable autonomous driving. The related research have been recognized by the Shenzhen Science and Technology Award, the CCF Award, and two Best Paper Awards at two international academic conferences. The key technologies have been applied by leading companies in autonomous driving.

The Trustworthy Software Research Center has been conducting research along the following four directions: 1) Software vulnerability analysis and detection techniques, 2) Software testing theories and techniques, 3) Automated software vulnerability repair theories and techniques, and 4) Modeling and simulation techniques for autonomous systems. The center has published more than 50 papers in international journals and conferences, applied for 10 patents, and secured 15 research projects at the national, provincial, municipal levels, as well as from leading IT companies. The developed technologies have been deployed and applied in companies such as Tencent, Huawei, and PopSquare.

The research team focuses on building a credible and interpretable model for heterogeneous data, integrating various human mobility data sources, solving issues of information scarcity and privacy protection in independent data sources. At the same time, the team combines machine learning theory with other domain knowledges to propose new theoretical methods and models to support in topics such as emergency management, transportation, energy, and epidemiology.