GAMES Webinar 2020 – 151期(海外专题) | Daniel Ritchie (Brown University)

【GAMES Webinar 2020-151期】(海外专题)

报告嘉宾:Daniel Ritchie (Brown University)


报告题目:Neurosymbolic 3D Models: Learning to Generate 3D Shape Programs


Daniel Ritchie is an Assistant Professor of Computer Science at Brown University, where he co-leads the Visual Computing group. His research sits at the intersection of computer graphics and machine learning: broadly speaking, he is interested in helping machines to understand the visual world, so that they can in turn help people to be more visually creative. His group’s current work focuses on data-driven methods for analyzing and synthesizing 3D scenes and the 3D objects that comprise them. His research is funded by gifts from Adobe, Autodesk, Pixar, and NVIDIA, as well as grants from DARPA and the National Science Foundation, including an NSF CAREER Award. He received his PhD from Stanford University and his undergraduate degree from UC Berkeley, both in Computer Science.


Generative models of 3D shapes promise compelling possibilities: the elimination of tedious manual 3D modeling in creative practices, powerful priors for autonomous vision and 3D reconstruction, and more. Procedural representations (such as probabilistic grammars or programs) are one such possibility: they offer high-quality and editable outputs but are difficult to author and often result in limited diversity among the output shapes. On the other extreme are deep generative models: given enough data, they can learn to generate any class of shapes, but their outputs exhibit artifacts and their representation is not easily interpretable or editable. In this talk, I’ll discuss the history of these two approaches and my group’s ongoing research agenda toward achieving the best of both worlds: neurosymbolic 3D models, a hybrid neural-procedural approach to 3D shape synthesis. Our work in this direction involves the design of a domain-specific language for defining the part structure of manufactured 3D shapes, as well as a deep generative model which can learn to write programs in this language for novel shape synthesis.



黄其兴是德克萨斯大学奥斯汀分校计算机科学的助理教授。他从斯坦福大学获得了计算机科学博士学位。在加入德克萨斯大学奥斯汀分校之前,他曾是芝加哥丰田技术学院的研究助理教授. 他的研究涉及计算机视觉、计算机图形学和机器学习领域。

GAMES主页的“使用教程”中有 “如何观看GAMES Webinar直播?”及“如何加入GAMES微信群?”的信息;

You may also like...