Xiang Gao

Xiang Gao is a Pre-tenure Associate Professor in the School of Software at Beihang University, China. His research focuses on applying program analysis, test generation, and formal methods to improve software quality through automated bug fixing and program synthesis. He has established significant collaborations with Fujitsu Laboratories of America, Microsoft Research, and other leading institutions in the software engineering field, demonstrating strong industry-academia connections.

Dr. Gao received his Bachelor's degree in Computer Science (Elite Class) from Shandong University in 2016, followed by a Ph.D. from the School of Computing at the National University of Singapore, where he also served as a Postdoctoral Fellow until December 2021. His educational background spans both Chinese and Singaporean academic institutions, providing him with a global perspective on software engineering research.

His primary research interests span multiple cutting-edge areas of software engineering:

• Program Analysis techniques for detecting and fixing software bugs with formal methods
• Software Security vulnerabilities with focus on automated repair methods
• Automated Program Repair systems that generate high-quality patches without overfitting
• Program Synthesis for creating transformation rules from examples
• Software Engineering for Artificial Intelligence (SE4AI) to improve AI model reliability and security
• Mobile Software Engineering with particular attention to UI testing and automation
• Deep Learning Security including model protection and obfuscation techniques

Dr. Gao's recent publication trajectory shows a strategic evolution toward integrating large language models with traditional software engineering approaches, particularly in test generation and program repair. His work on DNN modularization (NeMo, CNNSpliter, SeaM) represents an innovative approach to enhancing model reusability and security in resource-constrained mobile environments, addressing critical challenges in deploying AI on edge devices.

His scientific contributions have been recognized with multiple prestigious awards:

• ACM SIGSOFT Distinguished Paper Award for "ProveNFix: Temporal Property guided Program Repair" at FSE'24
• IEEE TCSE Distinguished Paper Award for "Investigating and Detecting Silent Bugs in PyTorch Programs" at SANER'24
• ACM SIGSOFT Distinguished Paper Award for "Modularizing while Training: A New Paradigm for Modularizing DNN Models" at ICSE'24
• Distinguished Artifact Award for "Automated Patch Backporting in Linux (Experience Paper)" at ISSTA'21

Dr. Gao actively mentors students at various levels, seeking "self-motivated Ph.D, master, undergraduate students and interns with strong programming skills" for his research projects. He serves on numerous program committees for top software engineering conferences including ICSE, ASE, ISSTA, and FSE, demonstrating his growing influence in the academic community. His research has been supported through collaborations with industry partners including Microsoft Research and Fujitsu Laboratories of America, translating theoretical advances into practical applications.

His laboratory focuses on several key research projects including Automated Software Vulnerability Repair (with techniques like Fix2Fit, VulnFix, and ExtractFix that address the overfitting problem in program repair), Program Synthesis for Program Transformation (including Semi-supervised synthesis and FixMorph for automated patch backporting in Linux), and Software Engineering for Artificial Intelligence (with projects like CNNSpliter, SeaM, and Sensei that apply software engineering principles to improve AI model usability and robustness). These projects represent cutting-edge work at the intersection of traditional software engineering and modern AI techniques, addressing critical challenges in software reliability and security.