报告题目:Real-Time Hair Simulation with Heptadiagonal Decomposition on Mass Spring System
报告人: 盛斌 副教授 上海交通大学
主持人: 何高奇 副教授
报告时间:2019年6月19日 周三15:00-17:00
报告地点:理科大楼B1002
报告摘要:
Simulating the detailed dynamics of hairs in real time is still a challenge problem in graphics due to the sheer number of strands and their complicated dynamics and self collisions. Previous interactive solutions either simplify the strand dynamics or reduce the DoFs in simulation at the cost of rich motion details. In this paper, we present a real-time simulation method for animating hair with full motion details. Our method is aimed at efficiently capturing the inextensibility, bending and torsion strand mechanics while presenting the stiction/repulsion and detailed collision effects in a real-time time budget. We utilize the local structure in the Selle mass-spring model and provide a fast implicit integration method based on heptadiagonal matrix decomposition in R3, which provides two to three orders of magnitude of acceleration to the original iterative-based approach like PCG. For efficiently capturing self interactions, we factorize the phenomenon into a coarse, globally coupled volumetric and a detailed collision view. The coarse behaviors are solved with an Eulerian method based on position based density control, while detailed collisions are efficiently handled with temporal coherent link update. Our method can faithfully simulate detailed hair motions interactively using desktop computer and is robust to various hair styles. We evaluate the efficiency and effectiveness of our method for simulating variant motions of the hair in different styles.
报告人简介:
盛斌,博士,上海交通大学计算机系计算机应用研究所副所长,博士生导师,副教授。 2011年获香港中文大学计算机科学与工程系博士学位,韩国成均馆大学客座教授, 上海市“浦江人才计划”获得者,图像处理领域著名 SCI 期 刊 IET Image Processing 的 Associate Editor。目前发表和录用百余篇学术期刊及会议论文,其中包括 Nature Communications, IEEE TVCG,IEEE VR, IEEE T. Image Processing, IEEE Transactions on Medical Imaging等在内的 SCI 期 刊 51 篇,以及MICCAI, IEEE ICCV, ECCV,IEEE VR 等重要会议论文 60 余篇, 学术专著 2部, 申请发明中国专利 34 项(授权6 项)。主持国家自然科学基金面上课题, 自然科学基金重点课题(子课题)和面上课题,国家 863 子课题等,2015年获上海市科技进步二等奖(第二完成人)。盛斌是中国计算机学会计算机辅助设计与图形学专委会委员,中国应用数学与工业设计学会几何设计专委会委员,人工智能学会智能 CAD 与数字娱乐专委会委员以及 IEEE 会员(Member),香港应用研究院(ASTRI)顾问。作为组委会主席举办中国计算机学会 CCF- YOCSE“智慧城市”论坛。相关研究成果获得 2015 年国际计算机图形学著名会议 Computer Graphics International 最佳论文奖,并展出于 2012 上海国际多媒体展和 2013 中国大屏幕展上,并指导学生在梅斯医疗创新大赛(2015)获全国二等奖。前期工作中已对基于深度学习的医学影像分析及辅助诊断、图像病灶的提取方法进行了研究(发表在IEEE TMI, MIA, MICCAI等)。