|
Ni Zhengyang I am a Ph.D. candidate at Xi'an Jiaotong University, advised by Prof. Fei Wang and Prof. Yu Guo at the Institute of Artificial Intelligence and Robotics. My research interests include 3D generation and embodied AI. I received my B.Eng. degree in Automation from Xi'an Jiaotong University in 2025. I work on generative and learning-based methods for 3D content creation and embodied AI. My goal is to move 3D models beyond virtual spaces by grounding them in the physical world. At the same time, I aim to enable robots not only to learn in high-fidelity 3D simulations, but also to transfer their learned capabilities to the real world, where they can perform diverse perception, interaction, and manipulation tasks in complex 3D environments. |
|
ResearchMy current research focuses on 3D generation and embodied AI. I am interested in models that can generate structured 3D content, reason about physical constraints, and support agents that perceive, plan, and interact in 3D environments. |
|
BrickAnything: Geometry-Conditioned Buildable Brick Generation with Structure-Aware Tokenization
Zhengyang Ni, Feng Yan, Yu Guo, Fei Wang arXiv, 2026 PDF / Code BrickAnything is a geometry-conditioned autoregressive framework for generating buildable brick structures from diverse 3D representations, using structure-aware tokenization to model local attachment relations and improve physical buildability. |
MiscellaneousThis website is adapted from Jon Barron's academic homepage template. |