会议介绍

Abs:
In our three-dimensional (3D) physical world, we constantly interact with 3D objects and environments. However, 3D generation introduces several key challenges. These include designing scalable 3D representations for high-resolution generative models, enabling creative generation beyond existing datasets, supporting intuitive 3D manipulation for flexible editing and control, and generating coherent scene-level environments that capture spatial relationships among multiple objects.
To address these challenges, this thesis focuses on four key directions: 3D representation, creative generation, human–AI manipulation, and scene-level generation. First, we develop scalable 3D representations that enable efficient training of high-resolution generative models. Second, we explore methods that allow generative models to produce creative and diverse 3D content beyond existing datasets. Third, we design interactive techniques that allow users to manipulate and customize 3D shapes using intuitive modalities such as text and images. Finally, we extend generative modeling from individual assets to coherent 3D scenes, enabling models to reason about spatial layouts and object interactions.
Our works advances the capability of AI-driven 3D creation and enabling more immersive digital experiences across applications such as simulation, content creation, and virtual or augmented reality.

Bio:
LIU Zhengzhe is an Assistant Professor in the School of Data Science at Lingnan University, Hong Kong. Before joining Lingnan University, he was a postdoctoral researcher at Carnegie Mellon University. He received his Ph.D. and M.Phil. degrees in Computer Science and Engineering from The Chinese University of Hong Kong (CUHK), supervised by Prof. FU Chi-Wing, and his B.Eng. degree from Shanghai Jiao Tong University. His research interests lie in computer graphics, computer vision, generative AI, with a focus on scalable 3D generation and world modeling. He has published 30+ papers in SIGGRAPH (Asia), TPAMI, TOG, CVPR, ICLR et al. and received MSRA Fellowship Nomination Award (33 in Asia).