题目: CAD-MLLM: Unifying Multimodality-Conditioned CAD Generation With MLLM
讲者: 高盛华, 香港大学
摘要: We introduce the CAD-MLLM, the first system capable of generating parametric CAD models conditioned on the multimodal input. Specifically, within the CAD-MLLM framework, we leverage the command sequences of CAD models and then employ advanced large language models (LLMs) to align the feature space across these diverse multi-modalities data and CAD models' vectorized representations. To facilitate the model training, we design a comprehensive data construction and annotation pipeline that equips each CAD model with corresponding multimodal data. Our resulting dataset, named Omni-CAD, is the first multimodal CAD dataset that contains textual description, multi-view images, points, and command sequence for each CAD model. It contains approximately 450K instances and their CAD construction sequences. To thoroughly evaluate the quality of our generated CAD models, we go beyond current evaluation metrics that focus on reconstruction quality by introducing additional metrics that assess topology quality and surface enclosure extent. Extensive experimental results demonstrate that CAD-MLLM significantly outperforms existing conditional generative methods and remains highly robust to noises and missing points.
个人简介: Shenghua Gao is an Associate Professor in the Department of Computer Science at the University of Hong Kong. Prior to joining HKU, he was a professor at ShanghaiTech University. His research interests include 3D reconstruction, image and video understanding and generation, 3D generation, AI4Science, etc. He has served as an area chair for over ten top conferences (CVPR, NeurIPS, ICCV, ACM MM, ECCV, etc.), and a publicity Chair for CVPR 2024. He also served as an associate editor for IEEE TPAMI, TMM, TCSVT, etc.
题目: DTGBrepGen: A Novel B-rep Generative Model through Decoupling Topology and Geometry
讲者: 陈发来, 中国科学技术大学
摘要: We introduce DTGBrepGen--a novel framework for automatically generating valid and high-quality Boundary Representation (B-rep) models, addressing the challenges posed by the complex interdependence between topology and geometry in CAD models. Unlike existing methods that prioritize geometric representation while neglecting topological constraints, DTGBrepGen explicitly models both aspects through a two-phase topology generation process followed by a Transformer-based diffusion model for geometry generation. Extensive experiments on diverse CAD datasets show that DTGBrepGen significantly outperforms existing methods in both topological validity and geometric accuracy, achieving higher validity rates and producing more diverse and realistic B-reps.
个人简介: 陈发来现为中国科学技术大学数学院教授、博士生导师,曾担任中国工业应用数学学会常务理事(2004-2016),中国工业应用数学学会几何设计与计算专委会主任(2015-2019),中国计算数学学会常务理事(2004—2012),安徽省数学会秘书长(2003-2019),安徽省工业与应用数学学会理事长(2023-), 国务院学位委员会数学学科评审组成员(2009-2019),教育部高等学校数学与统计学教学指导委员会委员(2006-2022),第十二、十三届国家自然科学基金委员会数学学科评委,《Computer Aided Geometric Design》,《CSIAM Transactions on Applied Mathematics》,《Journal of Computational Mathematics》,《计算机辅助设计与图形学学报》等期刊编委。曾于 1997年,2001年,2022 年三次获国家级教学成果二等奖。2001 年获教育部高校青年教师奖,2002年获国家自然科学基金杰出青年基金,2003 年获宝钢优秀教师奖特等奖,2008 年获中国计算机图形学杰出奖,2009 年获冯康科学计算奖,2010 年获全国百篇优博论文指导教师奖,2024 年获国际几何设计领域最高奖 John A. Gregory 纪念奖。 研究方向为计算机辅助几何设计与计算机图形学。近来感兴趣的研究课题包括:曲面隐式化的动曲面方法,T网格上的样条曲面,基于隐式曲面的三维散乱数据点重构、等几何分析及其在拓扑优化中的应用、基于稀疏优化的几何处理等。