BEGIN:VCALENDAR VERSION:2.0 PRODID:Linklings LLC BEGIN:VTIMEZONE TZID:America/Los_Angeles X-LIC-LOCATION:America/Los_Angeles BEGIN:DAYLIGHT TZOFFSETFROM:-0800 TZOFFSETTO:-0700 TZNAME:PDT DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:-0700 TZOFFSETTO:-0800 TZNAME:PST DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260805T205454Z LOCATION:Room 403 B DTSTART;TZID=America/Los_Angeles:20260721T104500 DTEND;TZID=America/Los_Angeles:20260721T123500 UID:siggraph_SIGGRAPH 2026_sess121@linklings.com SUMMARY:3D Generation DESCRIPTION:Nexus: Native Mesh Generation with Diffusion\n\nNexus is a rev olutionary two-stage diffusion framework for generating high-fidelity 3D p olygon meshes. By replacing slow sequential modeling with geometry structu re and topology diffusion, Nexus effortlessly handles complex, non-manifol d geometries and achieves more controllable and faster mesh gene...\n\n\nH anxiao Wang (Institute of Automation, Chinese Academy of Sciences; Univers ity of Chinese Academy of Sciences); Ying-Tian Liu and Yuan-Chen Guo (VAST ); Qi-Yuan Feng (CS Dept, Tsinghua University); Zi-Xin Zou and Ding Liang (VAST); Biao Zhang (Xi'an Jiaotong University); and Yan-Pei Cao (VAST)\n-- -------------------\nMatérn Noise for Triangulation-Agnostic Flow Matching on Meshes\n\nThis paper tackles the task of learning to generate signals over triangle meshes in a triangulation-agnostic manner, meaning the train ed model can be applied to different mesh triangulations effectively. It p roposes a specific noise distribution which is triangulation agnostic on m eshes, to be used ...\n\n\nTianshu Kuai (University of Montreal, Mila); Ar man Maesumi and Daniel Ritchie (Brown University); and Noam Aigerman (Univ ersity of Montreal, Mila)\n---------------------\nTempo3D: Efficient Tempo ral-Aware Fine-Tuning and Multi-View Latent Aggregation for 3D Generation\ n\nTempo3D is a resource-efficient framework for high-fidelity 3D generati on from a single image. By innovatively enhancing detail learning and reso lving single-view ambiguities without expensive retraining, it delivers st ate-of-the-art geometric fidelity. Additionally, Tempo3D empowers users wi th natu...\n\n\nHuizhi Zhu and Jiongming Qin (Wuhan University, School of Computer Science); Yusen Wang (DFRD, Dongfeng Motor Corporation Research&D evelopment Institute); and Chunxia Xiao (Wuhan University, School of Compu ter Science)\n---------------------\nPEGAsus: 3D Personalization of Geomet ry and Appearance\n\nPEGAsus enables personalized 3D shape generation by l earning reusable geometry and appearance concepts from a single reference shape, supporting both global and region-wise personalization. The learned concepts can be combined with text to synthesize novel personalized 3D as sets outperforming existi...\n\n\nJingyu Hu (The Chinese University of Hon g Kong), Bin Hu (The University of Hong Kong), Ka-Hei Hui (Autodesk Resear ch), Haipeng Li (The Hong Kong University of Science and Technology), Zhen gzhe Liu (Lingnan University), Daniel Cohen-Or (Tel Aviv University), and Chi-Wing Fu (The Chinese University of Hong Kong)\n---------------------\n AniGen: Unified $S^3$ Fields for Animatable 3D Asset Generation\n\nAniGen generates animatable 3D assets from a single image by jointly producing ge ometry, skeletons, and skinning weights. Its unified S³ field representati on enables consistent, high-quality rigged asset creation across diverse c ategories, including animals, humans, cartoon characters, and articulat... \n\n\nYi-Hua Huang (The University of Hong Kong (HKU)), Zi-Xin Zou (VAST), Yuting He (The Chinese University of Hong Kong), Chirui Chang (The Univer sity of Hong Kong (HKU)), Cheng-Feng Pu (Tsinghua University), Ziyi Yang ( The University of Hong Kong (HKU)), Yuan-Chen Guo and Yan-Pei Cao (VAST), and Xiaojuan Qi (The University of Hong Kong (HKU))\n--------------------- \nMeshFlow: Mesh Generation with Equivariant Flow Matching\n\nMeshFlow is a powerful diffusion-based framework utilizing equivariant modeling to tra nsform random triangle soups into high-quality meshes in less than one sec ond.\n\n\nQi Sun (City University of Hong Kong), Kiyohiro Nakayama (Stanfo rd University), Jing Yan (Cornell Tech), Qixing Huang (The University of T exas at Austin), Alexander Rush (Cornell Tech), Leonidas Guibas and Gordon Wetzstein (Stanford University), Jing Liao (City University of Hong Kong) , and Guandao Yang (The University of Texas at Austin)\n------------------ ---\nStrips as Tokens: Artist Mesh Generation with Native UV Segmentation\ n\nWe propose Strips as Tokens (SATO), a novel framework with a token orde ring strategy inspired by triangle strips. By constructing the sequence as a connected chain of faces that explicitly encodes UV boundaries, our met hod naturally preserves the organized edge flow and semantic layout charac teristi...\n\n\nRui Xu and Dafei Qin (University of Hong Kong); Kaichun Qi ao (ShanghaiTech University); Qiujie Dong and Huaijin Pi (University of Ho ng Kong); Qixuan Zhang, Longwen Zhang, Lan Xu, and Jingyi Yu (ShanghaiTech University); Wenping Wang (Texas A&M University); and Taku Komura (Univer sity of Hong Kong)\n---------------------\n3D Generation - Interactive Dis cussion\n\nAfter the summary presentations, attendees will participate in an interactive discussion. Outside the room will be a series of poster boa rds for authors to gather around with the audience. Authors are invited to bring any material related to their paper that could instigate further co nversation such...\n\n\nInterest Area: Research & Education\n\nKeyword: Ar tificial Intelligence/Machine Learning, Geometry, Modeling\n\nRegistration Category: Full Conference Supporter, Full Conference\n\nSession Chair: Mi rela Ben-Chen (Technion – Israel Institute of Technology) END:VEVENT END:VCALENDAR