I am Mingfei Cheng (程明飞), a third-year PhD candidate at Singapore Management University (SMU, 2022-Current), supervised by Prof. Xiaofei Xie. Prior to joining SMU, I obtained my M.S. and B.S. degrees from the Beijing University of Posts and Telecommunications (BUPT) and Queen Mary University of London (QMUL), under the advisement of Prof. Kaili Zhao and Prof. Yajing Xu.
My current research mainly focuses on Software Engineering for Autonomous Driving Systems (SE4ADS) and Artificial Intelligence for Software Engineering (AI4SE). Additionally, I remain interested in my previous research on Computer Vision (CV), as well as topics in Natural Language Processing (NLP).
I am also focusing on developing open-source tools that are easy to understand and use for the SE4ADS community. I warmly welcome anyone interested to connect and collaborate with me 😁.
🔥 News
- 2024.11: 🎉🎉 One paper is accepted by ICSE 2025!
📖 Selected Publications (Full List)
Decictor: Towards Evaluating the Robustness of Decision-Making in Autonomous Driving Systems
Mingfei Cheng, Xiaofei Xie, Yuan Zhou, Junjie Wang, Guozhu Meng, Kairui Yang
ICSE 2025, Research Track, Pre-print, Code will be released soon
This paper is the first to evaluate the robustness of decision optimality in Autonomous Driving Systems.
BehAVExplor: Behavior Diversity Guided Testing for Autonomous Driving Systems
Mingfei Cheng, Yuan Zhou, Xiaofei Xie
ISSTA 2023, Technical Track, Paper, Project Code
This paper is the first to leverage ego behavior as feedback to enhance the diversity of generated scenarios in ADS testing, thereby uncovering diverse intrinsic weaknesses within the ADS.
Joint Topology-preserving and Feature-refinement Network for Curvilinear Structure Segmentation
Mingfei Cheng, Kaili Zhao, Xuhong Guom, Yajing Xu, Jun Guo
ICCV 2021, Main Conference, Paper, Project Code
This paper proposes a novel framework to jointly optimize topology preservation and pixel-level performance for curvilinear structure segmentation.
🎯 Leading Projects (Full List)
I have developed several tools for ADS testing, some of which are not yet publicly released (with plans to release them by the end of 2024). All leading projects were independently developed by me and may not be absolutely perfect. I warmly welcome anyone interested in helping to maintain and improve these projects.
- DriveTester (Github Link)
I designed and implemented a unified testing platform for Autonomous Driving Systems (ADS) testing, which integrates my previous projects, including ApolloSim. This platform aims to provide various testing techniques (e.g., AVFuzzer) for quick adoption in research. This is a long-term, ongoing project, I welcome anyone interested to join and engage in discussions. - ApolloSim
I developed a simple, research-driven simulation environment for testing the decision-making components of Baidu Apollo, including Prediction, Planning, Routing, and Control. This platform does not require a graphics rendering engine, making it easy to use on Linux servers without a GUI. - CarlaFuzz
I implemented an ADS testing tool based on CARLA, which incoroperates various ADSs and existing ADS testing techniques. This project will be released after the acceptance of my new papers.
🏆 Awards
- 2023.06 ACM SIGSOFT CAPS Travel Funds.
- 2023.03 The 3rd Place Winner in WSDM Ideation Challenge.
- 2019.06 Outstanding Graduates in Beijing.
- 2016-2018 The First Level Presidential Scholarship from Beijing University of Posts and Telecommunications.
📝 Services
Conference
- PC member of ICSE 2025, Artifact Evaluation
- PC member of ICSE 2025, SE4ADS Track
- Junior PC member of MSR 2025, Technical Track
- PC member of ISSTA 2024, Artifact Evaluation
- Junior PC member of MSR 2024, Technical Track
- PC member of USENIX Security ‘24, Artifact Evaluation
- Reviewer of ACCV 2022 and ECCV 2022
Journal
- IEEE Transactions on Software Engineering
- IEEE Transactions on Medical Imaging
- Engineering Applications of Artificial Intelligence
- Neurocomputing
💻 Internships
- 2021.07 - 2022.01, Alibaba Cloud, Computer Vision Algorithm Intern, Hangzhou & Beijing, China.
- 2021.04 - 2021.07, Beijing Research Institute of China Telecom, Computer Vision Algorithm Intern, Beijing, China.