SLAI Seminar 17th Session will be discussing the topic on "Hunyuan 3D: From Object Modeling to 3D World Generation", from 10am to 11:30am, December 22th (Monday) at B411 Lecture Hall, online participation is welcome (Tencent Meeting ID: 312-512-332)

报告主题：混元3D-从物体建模到3D世界生成 

时间：2025年12月22日（周一）上午10点-11点30

地点: 深圳河套学院B411阶梯教室

线上参与：腾讯会议号 312-512-332

About the Speaker:

Dr. Chunchao Guo is a Principle Research Scientist at Tencent and the Head of Hunyuan 3D. His research interests focus on 3D generative large models and world models. He has authored and co-authored over 30 peer-reviewed papers in prestigious journals and top-tier conferences. Additionally, he has secured 20 championships in international technical competitions, and during his doctoral studies, he was awarded the Outstanding Paper Award at the National Conference on Image and Graphics (NCIG).Dr. Guo has spearheaded his team in developing one of the industry’s pioneering 3D-DiT (3D Denoising Diffusion Implicit Models) large models, which have been deployed across diverse scenarios such as gaming and social media platforms. He also oversaw the release of HunyuanWorld—the first open-source 3D world model of its kind—garnering extensive recognition and attention within the academic and industrial communities.

郭春超博士 腾讯专家研究员、混元3D负责人，先后从事广告AI、生成式大模型研究与落地工作，已在CVPR等各类期刊会议发表论文三十余篇，在国内外权威技术竞赛获冠军20项，博士期间获得全国图像图形学术会议优秀论文奖。带领团队研发业界最早一批3D-DiT大模型，应用于游戏、地图、社交等多个场景。发布首个开源3D世界模型HunyuanWorld，在业界获得广泛关注。

Abstract：

We live in a three-dimensional world and perceive the physical reality world through tangible 3D objects. With the rapid development of generation models, 3D AIGC has emerged as a cutting-edge research focus, widely regarded by the industry as the next frontier poised for groundbreaking innovations following text, image, and video modalities. Meanwhile, the rise of embodied intelligence and world models has further fueled attention and discussions surrounding 3D generation. This presentation outlines the current Research and Development Roadmap of Hunyuan 3D and world model, as well as the latest progress and technical outlook.

我们身处于三维世界，通过实实在在的三维物体感知物理世界。伴随着生成式大模型技术发展，3D AIGC也快速成为了新的研究焦点，被业界认为是继文本、图片、视频模态之后有望实现重大突破的领域。同时，随着具身智能、世界模型的兴起，3D生成也获得更多关注和讨论。本次报告介绍当前混元3D的布局、世界模型的研发进展与技术展望。