GAMES Webinar 2023 – 289期(精细几何建模与优化) | 王逸群(重庆大学),沈天畅(多伦多大学,英伟达)
【GAMES Webinar 2023-289期】(几何专题-精细几何建模与优化)
报告嘉宾:王逸群(重庆大学)
报告时间:2023年8月24号星期四晚上8:00-8:30(北京时间)
报告题目:基于多视图的高质量神经隐式表面重建方法
报告摘要:
图像驱动的三维曲面重建是计算机图形学和视觉领域的核心挑战之一。开创性的NeRF方法通过对体积中每个点的密度函数和与视角相关的颜色进行建模,为后续研究提供了新的思路。许多研究工作将基于神经渲染的方法应用于隐式表面模型,从而在无需三维监督信息的情况下实现表面重建。在本次报告中,我们将介绍我们在神经隐式表面重建方面的最新进展。这些进展涵盖了如何利用高频信息重建隐式曲面的细节,如何引入三平面表示优化表面重建过程,以及针对具有光泽物体表面的重建方法。
讲者简介:
王逸群,重庆大学副教授,硕士生导师,博士毕业于中国科学院自动化研究所,本科毕业于重庆大学。2021-2022在阿卜杜拉国王科技大学(KAUST)担任博士后研究员。主要研究方向为计算机图形学、几何处理与学习、三维重建。在SIGGRAPH,NeurIPS,CVPR,AAAI,TOG,T-PAMI,TVCG,TIP等会议与期刊(CCF A 类)上发表论文11篇。曾获得中科院院长奖、重庆市优秀毕业生等奖项。担任中国计算机学会(CCF)计算机辅助设计与图形学专业委员会执行委员,以及NeurIPS、CVPR、ICCV、ECCV、TVCG等会议或期刊的审稿人。先后主持了国自然青年、重庆市面上,负责重庆市科技创新重大研发项目子课题、国自然重点项目子课题等,部分研发技术已应用于相关公司。
讲者主页:https://sites.google.com/view/yiqun-wang/home
报告嘉宾:沈天畅(多伦多大学,英伟达)
报告时间:2023年8月24号星期四晚上8:30-9:00(北京时间)
报告题目:Flexible Isosurface Extraction for Gradient-Based Mesh Optimization
报告摘要:
This work considers gradient-based mesh optimization, where we iteratively optimize for a 3D surface mesh by representing it as the isosurface of a scalar field, an increasingly common paradigm in applications including photogrammetry, generative modeling, and inverse physics. Existing implementations adapt classic isosurface extraction algorithms like Marching Cubes or Dual Contouring; these techniques were designed to extract meshes from fixed, known fields, and in the optimization setting they lack the degrees of freedom to represent high-quality feature preserving meshes, or suffer from numerical instabilities. We introduce FlexiCubes, an isosurface representation specifically designed for optimizing an unknown mesh with respect to geometric, visual, or even physical objectives. Our main insight is to introduce additional carefully-chosen parameters into the representation, which allow local flexible adjustments to the extracted mesh geometry and connectivity. Extensive experiments validate FlexiCubes on both synthetic benchmarks and real-world applications, showing that it offers significant improvements in mesh quality and geometric fidelity.
讲者简介:
Tianchang Shen is a PhD student at the University of Toronto, advised by Professor Sanja Fidler. He also serves as a research scientist at the NVIDIA Toronto AI lab. His research aims to enhance 3D reconstruction and generation using AI techniques, leveraging innovative 3D representations. His work has been published at venues such as SIGGRAPH, NeurIPS, CVPR, and ECCV. Additionally, his contributions have been integrated into NVIDIA products, including Picasso3D, Neural DriveSim, and 3D MoMa.
讲者主页:http://www.cs.toronto.edu/~shenti11/
主持人简介:
郭建伟,中科院自动化所模式识别国家重点实验室副研究员,硕士生导师,中科院青促会会员。主要研究方向为三维几何建模、形状分析及特征表示学习,在主流期刊或会议上发表论文60余篇,包括国际期刊ACM TOG、IEEE TVCG、IEEE TIP、CAD、IEEE TMM及国际会议SIGGRAPH/Asia、CVPR、ECCV等,授权发明专利17项。曾获得中国体视学学会青年科学技术奖、陆增镛CAD&CG高科技二等奖、3次国内外会议/期刊最佳论文奖/提名奖、中国图学学会科技进步二等奖、CCF-腾讯犀牛鸟科研基金优秀奖等。担任SCI期刊the Visual Computer、IET Image Processing编委,以及Visual Computing for Industry, Biomedicine, and Art期刊青年编委。
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;
GAMES主页的“资源分享”有往届的直播讲座的视频及PPT等。
观看直播的链接:http://webinar.games-cn.org