Quality and reliability of AI systems currently become a pressing concern, where many industry sectors demand a certain level of quality guarantee so that the AI systems can be eventually adopted, especially for safety- and mission-critical domains. In practice, an AI system can often be rather complex in nature, spanning AI models, traditional software components, and their interactions. Lei Ma’s research focuses on providing both fundamental quality assurance methodology and systematic engineering support for building complex AI systems to make them more reliable, safe and secure, bridging the gap between AI and its real-world applications. He is continuously proposing novel techniques of quality assessment, quality issue detection, localization, root cause analysis, repairing, etc., specially designed for both AI models at the unit level and AI systems at the whole system level, to address the AI quality concerns across domains, e.g., autonomous driving, video games, DeepFake, healthcare, etc.
- ACM SIGSOFT Distinguished Paper Award, The 34th IEEE/ACM Conference on Automated Software Engineering (ASE 2019), Nov. 2019
- Baidu-NASAC Academic Star Award, China Computer Federation Technical Committee on Software Engineering (CCF TCSE), Nov. 2018
- ACM SIGSOFT Distinguished Paper Award, The 33rd IEEE/ACM International Conference on Automated Software Engineering (ASE 2018), Sep. 2018
- ACM SIGSOFT Distinguished Paper Award, The 30th IEEE/ACM International Conference on Automated Software Engineering (ASE 2015), Nov. 2015
- Championship & Best Testing Tool Award, 2015 International Search-Based Software Testing Competition, May 2015
- Guo, Q., Juefei-Xu, F., Xie, X., Ma, L., Wang, J., Yu, B., ... & Liu, Y. (2020). Watch out! motion is blurring the vision of your deep neural networks. Advances in Neural Information Processing Systems, 33.
- Peng, Z., Yang, J., Chen, T. H., & Ma, L. (2020, November). A first look at the integration of machine learning models in complex autonomous driving systems: a case study on Apollo. In Proceedings of the 28th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering (pp. 1240-1250).
- Du, X., Xie, X., Li, Y., Ma, L., Liu, Y., & Zhao, J. (2019, August). Deepstellar: Model-based quantitative analysis of stateful deep learning systems. In Proceedings of the 2019 27th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering (pp. 477-487).
- Ma, L., Juefei-Xu, F., Zhang, F., Sun, J., Xue, M., Li, B., ... & Wang, Y. (2018, September). Deepgauge: Multi-granularity testing criteria for deep learning systems. In Proceedings of the 33rd ACM/IEEE International Conference on Automated Software Engineering (pp. 120-131).
- Ma, L., Artho, C., Zhang, C., Sato, H., Gmeiner, J., & Ramler, R. (2015, November). Grt: Program-analysis-guided random testing (t). In 2015 30th IEEE/ACM International Conference on Automated Software Engineering (ASE) (pp. 212-223). IEEE.
CIFAR is a registered charitable organization supported by the governments of Canada, Alberta, Ontario, and Quebec as well as foundations, individuals, corporations, and international partner organizations.