When in high school, Xinyu Lei discovered a keen interest in computer programming, inspiring him to major in computer science as an undergraduate. Gradually, he gained an intensive interest in computer science research. Following completion of his Ph.D. in computer science at Michigan State University, he says the natural choice was to become a professor and computer scientist.
Lei is an assistant professor in Michigan Tech’s Department of Computer Science. His cybersecurity research is applied to help users protect their private information while using IoT devices and machine learning algorithms. His current projects focus on machine learning security.
For instance, he is studying a deep learning technique that can be used to generate faked face images—often used for malicious purposes—with the purpose of learning to detect them.
Some examples of faked face images generated by this deep learning technique are found here. Note that refreshing the webpage reveals a new faked image.
During his Ph.D. studies, Lei also developed techniques to enable an adversary to decrypt a video taken by a smart camera, again with the aim of preventing such threats. Find more details about this project here.
Looking forward in his research, Lei is excited about a recently proposed deep learning technique named CycleGAN, which can be used for image style transformation. For example, given a picture taken in the summer,CycleGAN can automatically generate a picture taken in the winter—using the same content. Find more details about CycleGAN here.
Lei teaches and mentors in a student-centric manner. “First, I study to understand what students need and what industry and academia need, and I develop my course accordingly,” he says. Then, through the completion of Lei’s carefully prepared homework and project assignments, students can gain a deep understanding of the subject being studied. This semester Lei is teaching Computer Security (CS 4471/5471).
Lei is seeking multiple Ph.D. students with interests in machine learning or cybersecurity research. Enrollment can begin in Spring 2022, Fall 2022, Spring 2023, Fall 2023. Learn more on Lei’s faculty webpage.
Lei enjoys sharing his specialized knowledge to offer a scientific take on popular contemporary topics, such as Bitcoin and Meta, in conversation with family and friends. He is a sports fan, and he likes to play tennis, soccer, and volleyball, watch movies, and stay current with the news.