I am a postdoctoral researcher at UCL and an incoming Assistant Professor and Chancellor’s Fellow at the University of Strathclyde. My research focuses on developing robust, efficient, and trustworthy AI systems, with applications spanning computer vision, adversarial robustness, and learning with imperfect information.

The Hong Kong Laureate Forum represents a rare and inspiring platform to engage with world-class scientists whose work forms the theoretical bedrock of my field. I am especially excited by the Forum’s interdisciplinary vision, where mathematics, physics, and life sciences intersect with AI—a perspective that will directly inform my future research agenda and academic leadership.

As I transition into a faculty role, I am committed to fostering responsible AI development and mentoring the next generation of researchers. Participating in the Forum will not only broaden my intellectual horizon, but also allow me to exchange ideas and build collaborations with other rising talents across the globe. It will be a unique opportunity to be inspired by the insights, values, and scientific journeys of the laureates, and to carry that inspiration into my academic career.

Queen Mary University of London

University of Trento

ChenFeng-Oct2023.pdf

My research focuses on building robust, efficient, and trustworthy machine learning systems that can operate reliably under real-world uncertainty. I investigate learning with imperfect data—such as noisy labels, weak supervision, and out-of-distribution samples—by designing principled frameworks rooted in statistical theory and scalable optimization.

During my PhD and postdoctoral training, I developed methods in self-supervised and contrastive learning, provable safety certification under adversarial attacks, and novel noise-robust algorithms. These contributions were published in top-tier venues including CVPR, AAAI, and ACM MM. I am particularly interested in bridging the gap between theoretical insights and practical deployment, ensuring AI systems are not only accurate but also safe, interpretable, and socially responsible.

Currently, I am a Leverhulme Research Fellow at University College London and will soon join the University of Strathclyde as Assistant Professor and Chancellor’s Fellow. My future research will extend to general AI safety, interdisciplinary applications in healthcare and law, and the development of AI systems that can explain and adapt to human expectations. I aim to foster international collaboration and cross-domain dialogue to shape the next generation of responsible AI technologies.