特邀报告3：Hao Zhang,Learning Generative Models of 3D Shapes

简介

Unlike images and video, 3D shapes are not confined to one standard representation. This is one of the challenges we face when developing deep neural networks (DNNs) to learn generative models of 3D shapes or virtual scenes. So far, voxel grids, multi-view images, point clouds, and integrated surface patches have all been considered. In this talk, I show that traditional convolutional neural networks operating on pixels/voxels may not be best suited for the task. I first present IM-Net, our recent work on learning implicit functions, and show the innate ability of implicit models for shape generation and 3D reconstruction. Interestingly, a small tweak of IM-NET, with an added branching layer, would allow the new network, BAE-Net, to perform unsupervised shape segmentation. Furthermore, by explicitly exploiting the interpretability of BAE-Net, we arrive at BSP-Net (BSP for binary space partitioning), a DNN which directly outputs compact, structured meshes, with sharp features. While the IM-BAE-BSP-Net series exemplify the power of "deep shape priors" for part extraction, these works do not learn part relations. In the second part of my talk, I present a generative recursive autoencoder to learn shape structures in the form of attributed trees. Such an approach decouples coarse and fine-grained learning of structured data, which is applicable to both 3D shapes, indoor scenes, and digital documents.

直播
日程
时间：10月24日8:30-9:20
地点：报告厅
嘉宾

Hao (Richard) Zhang

演讲摘要：

个人简介：Hao (Richard) Zhang is a professor in the School of Computing Science at Simon Fraser University (SFU), Canada, where he directs the graphics (GrUVi) lab. He obtained his Ph.D. from the University of Toronto, and MMath and BMath degrees from the University of Waterloo, all in computer science. His research is in computer graphics with special interests in geometric modeling, shape analysis, 3D content creation, machine learning, and computational design and fabrication, and he has published more than 120 papers on these topics. Richard is an associate editor-in-chief for IEEE CG&A, a past editor-in-chief for Computer Graphics Forum, and an associate editor for IEEE TVCG. He has served on the program committees of all major computer graphics conferences and is SIGGRAPH Asia 2014 course chair, a paper co-chair for SGP 2013, GI 2015, and CGI 2018, and a program chair for International Geometry Summit 2019. Awards won by Richard include an NSERC DAS (Discovery accelerator Supplement) Award in 2014, Best Paper Awards from SGP 2008 and CAD/Graphics 2017, a Faculty of Applied Sciences (FAS) Research Excellence Award at SFU in 2014, a National Science Foundation of China (NSFC) Overseas Outstanding Young Researcher Award in 2015, and a Google Faculty Award in 2020. Richard has been a visitor professor at Stanford University, Shenzhen University, and the Beijing Film Academy. Hao (Richard) Zhang