About Me

Hello, I am Chucheng Xiang, a master's student in Computational Mathematics at the University of Science and Technology of China (USTC). I received my B.S. degree in the same major from USTC in 2024 and expect to graduate in 2027. I am advised by Prof. Ligang Liu and work closely with Dr. Zhongyuan Liu and Prof. Ziqi Wang.

My research interests lie in multimodal foundation models, LLM training, and agentic systems, with a focus on 3D perception, interaction, and generation.

News

Oct 2026
Joined Tencent Hunyuan (VLM team) as a Research Intern, working on improving the spatial perception of multimodal foundation models.
Oct 2026
Awarded the National Scholarship for the second time.
Jul 2026
Won first place in the CAD Wireframe Neural Compression Challenge at CSIAM GDC 2026, sponsored by MathMagic.
Jul 2026
Two papers, GALA: Geometry-Aware Language Model for Controllable Object Arrangement and PART: Learning 3D Part Assembly and Retrieval with Transformers, were accepted to SIGGRAPH Asia 2026.
Jun 2026
Our paper NaLA: A 3D Native LLM Layout Agent for High-quality 3D Scene Generation was accepted to ECCV 2026.
Nov 2025
Our paper Co-Layout: LLM-driven Co-optimization for Interior Layout was accepted to AAAI 2026.
Oct 2025
Awarded the National Scholarship.
Aug 2025
Joined Tencent IEG as a Research Intern working on 3D perception, scene generation, and large language models.
Jul 2025
Our paper Scene Layout via Conceptual Design was accepted to SMI 2025 and will be published in Computers & Graphics.

Publications

GALA: Geometry-Aware Language Model for Controllable Object Arrangement Chucheng Xiang, Runze Wang, Zhi Deng, Ruchao Bao, Yuanwei Zhang, Yuxuan Xie, Hanliu Wang, Liangzhen Fei, Wenzheng Wu, Cheng Wan, Peifeng Li, Zhongyuan Liu, Ligang Liu
The 19th ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia (SIGGRAPH Asia 2026).

paper doi project page code model dataset

PART: Learning 3D Part Assembly and Retrieval with Transformers Ruchao Bao, Wenzheng Wu, Chucheng Xiang, Zhongyuan Liu, Yuan Liu, Jinxin Dong, Ligang Liu, Ziqi Wang
The 19th ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia (SIGGRAPH Asia 2026).

project page

NaLA: A 3D Native LLM Layout Agent for High-quality 3D Scene Generation Cheng Wan, Yongsen Mao, Wenzheng Wu, Yuxuan Xie, Chucheng Xiang, Runze Wang, Monica Zhang, Zhongyuan Liu, Rushi Dai, Yuan Liu
The 19th European Conference on Computer Vision (ECCV 2026). (Poster)

paper project page code

Co-Layout: LLM-driven Co-optimization for Interior Layout Chucheng Xiang, Ruchao Bao, Biyin Feng, Wenzheng Wu, Zhongyuan Liu, Yirui Guan, Ligang Liu
The 40th Annual AAAI Conference on Artificial Intelligence (AAAI 2026), Vol. 40, No. 17, pp. 14371–14379. (Poster)

paper doi project page code

Scene Layout via Conceptual Design Wenzheng Wu, Chucheng Xiang, Zhi Lin, Yirui Guan, Ruchao Bao, Zhongyuan Liu, Ziqi Wang, Ligang Liu
Shape Modeling International (SMI 2025), published in Computers & Graphics, Vol. 136, Article 104553.

paper pdf project page

Honors & Awards

First Place, CAD Wireframe Neural Compression Challenge, CSIAM GDC 2026 (MathMagic) 2026
National Scholarship 2025, 2026
Chiang Chen Industrial Charity Foundation Scholarship 2025
First-Class Academic Scholarship for Master's Students 2024-2026
Outstanding Undergraduate Graduate, USTC 2024
Outstanding Student Scholarship, USTC 2021-2023
Provincial First Prize, National Undergraduate Mathematics Competition (Mathematics A) 2022

Education & Experience

Tencent, Hunyuan VLM

HY4 Post-training: Built training data from 20K+ 3D scenes and a physics- and vision-based automated evaluation system, improving the model's spatial perception, scene generation, and Vibe-Gaming Agent capabilities.

Tencent, IEG LightSpeed Studios

Tackled fine-grained object placement in game scene generation: designed a geometry-aware controllable placement algorithm, trained a 3D multimodal model, and built a high-quality scene dataset and evaluation protocol. The work became the GALA paper (SIGGRAPH Asia 2026) and won the IEG Tech Future Project Excellence Award.