Hello! I’m Kewei, currently a Master’s student at The University of Hong Kong. Prior to that, I received my Bachelor’s degree from The Chinese University of Hong Kong, Shenzhen. My research mainly concentrates on 3D vision and graphics. I currently work on topics about 3D AIGC including 3D neural representations, 3D generative models, and 3D-aware video generation.

Research Focus

• Generative Models (2D & 3D)

  Designing and training models that can synthesize realistic images, point clouds, and meshes—bridging the gap between data-driven deep learning and traditional graphics pipelines.

• Shape Representation

  Investigating compact, expressive ways to encode geometry (e.g., implicit functions, signed distance fields, neural radiance fields) so that 3D objects can be stored, manipulated, and rendered with high fidelity.

• Deep Learning Methods for numerical PDE solvers

  Applying finite-difference, finite-element, and spectral techniques to solve partial differential equations arising in physics-based simulation, geometry processing, and volumetric reconstruction.

Let’s Connect

I thrive on interdisciplinary collaboration and love discussing novel ideas—whether it’s a new architecture for mesh generation, a faster solver for time-dependent PDEs.
Feel free to reach out via email: kewei.shi@connect.hku.hk