Tianye Ding

Hi! I am currently a Research Assistant at Visual Intelligence Lab at Northeastern supervised by Prof. Huaizu Jiang. I graduated magna cum laude with a B.S. Computer Science degree from Khoury College of Computer Sciences, Northeastern University.

My research interests lie at the intersection of computer vision and geometric deep learning. I am driven by the goal of bridging the gap between interpretable classical methods and modern black-box models. Specifically, I explore how integrating explicit structural priors—such as geometry, compositionality, and physical laws—can unlock capabilities beyond those accessible through scaling alone. My work spans from designing efficient, mathematically grounded perception algorithms to developing 3D and 4D generative models that maintain physical consistency across time and space.

One of my personal passions is crafting blog posts that highlight the key insights and my thought processes behind the innovative solutions that drive breakthroughs in my research projects. I firmly believe that sharing these experiences can not only inspire others but also shed light on the fundamental issues at the core of the topics I’ve explored.