Zekun's research focuses on controllable generative models, particularly for synthesizing photorealistic 3D worlds. His works include user-guided video generation, intuitive 3D shape manipulation with learned priors, and more recently photorealistic 3D world synthesis guided by Minecraft-like input. His research goal is to develop algorithms that learn from real-world visual data and then apply the knowledge to helping human creators build their own photorealistic 3D worlds effortlessly.
Zekun is a 4th year PhD student in computer science at Cornell Tech and Cornell University, working with Professor Serge Belongie. His research interests are machine learning and its applications to computer vision and computer graphics. Prior to this, he received his B.S. in computer science from Beihang University, China.