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:20260805T205456Z
LOCATION:Room 408 B
DTSTART;TZID=America/Los_Angeles:20260723T154500
DTEND;TZID=America/Los_Angeles:20260723T171500
UID:siggraph_SIGGRAPH 2026_sess138@linklings.com
SUMMARY:Gaussian Reconstruction
DESCRIPTION:GeoQuery: Geometry-Query Diffusion for Sparse-View Reconstruct
 ion\n\nGeoQuery is a geometry-guided single-step diffusion framework for m
 itigating query contamination in existing multi-view self attention mechan
 ism. By retrieving proxy features from clean reference views via geometric
  correspondences, it bypasses corrupted query features and enables high-fi
 delity rest...\n\n\nXiao Cao (University of Electronic Science and Technol
 ogy of China), Yuze Li (Tianjin University), Youmin Zhang and Jiayu Song (
 Rawmantic AI), Cheng Yan (Tianjin University), and Wen Li and Lixin Duan (
 University of Electronic Science and Technology of China)\n---------------
 ------\nVidSplat: Gaussian Splatting Reconstruction with Geometry-Guided V
 ideo Diffusion Priors\n\nWe present VidSplat, a generative reconstruction 
 framework that leverages powerful video diffusion priors to synthesize nov
 el views that compensate for missing input coverage, and thereby recover c
 omplete 3D scenes from sparse inputs. VidSplat performs robustly to sparse
  input and even a single imag...\n\n\nJimin Tang, Wenyuan Zhang, Junsheng 
 Zhou, and Zian Huang (Tsinghua University); Kanle Shi and Shenkun Xu (Kuai
 shou Technology); Yu-Shen Liu (Tsinghua University); and Zhizhong Han (Way
 ne State University)\n---------------------\nImmediate 3D Gaussian Splat R
 econstruction of Unordered Input with Global Consistency\n\nWe propose a 3
 D Gaussian Splatting reconstruction method with immediate feedback that ha
 ndles unordered image captures and very large scenes. Through fast place-r
 ecognition-driven matching, cluster-based loop closure with graph-propagat
 ed correction, and a progressive Gaussian hierarchy construction...\n\n\nA
 ndreas Meuleman (Inria, Université de Rennes); Linus Franke (Inria, Univer
 sité Côte d'Azur); Boris Zhestiankin (Inria, Université Côte d'Azur; EPFL)
 ; Camille Montemagni (Inria, Université de Rennes); and George Drettakis (
 Inria Université Côte d'Azur)\n---------------------\nMoonSplat: Monocular
  Online Gaussian Splatting with Sim(3) Global Optimization\n\nMoonSplat is
  a robust and efficient online voxelized 3DGS reconstruction framework int
 egrated with global Sim(3) optimization, which enables reliable camera tra
 cking and efficient global loop closure for both camera poses and voxelize
 d 3DGS. MoonSplat is equipped with a color residual learning strat...\n\n\
 nGuo Pu and Yixuan Han (Peking University); Haofeng Li, Yao Zhang, and Hui
  Zhou (Beijing Hydrogen Intelligent Tech. Co., Ltd.); and Zhouhui Lian (Pe
 king University)\n---------------------\nRaDe-GS: Rasterizing Depth in Gau
 ssian Splatting\n\nGaussian Splatting enables high-quality, real-time nove
 l-view synthesis, but extracting detailed 3D geometry remains challenging.
  We introduce a rasterized approach to render the depth and normal maps fo
 r general 3D Gaussian primitives. Our method significantly enhances shape 
 reconstruction accuracy...\n\n\nBaowen Zhang and Chuan Fang (Hong Kong Uni
 versity of Science and Technology), Rakesh Shrestha (Simon Fraser Universi
 ty), Yixun Liang (Hong Kong University of Science and Technology), xiaoxia
 o Long (Nanjing University), Ping Tan (Hong Kong University of Science and
  Technology), and Baowen Zhang\n---------------------\nGaussian Reconstruc
 tion - Interactive Discussion\n\nAfter the summary presentations, attendee
 s will participate in an interactive discussion. Outside the room will be 
 a series of poster boards for authors to gather around with the audience. 
 Authors are invited to bring any material related to their paper that coul
 d instigate further conversation such...\n\n\nInterest Area: Research & Ed
 ucation\n\nKeyword: Generative AI, Geometry, Modeling, Rendering\n\nRegist
 ration Category: Full Conference Supporter, Full Conference\n\nSession Cha
 ir: Hongzhi Wu (Zhejiang University; State Key Laboratory of CAD&CG, Zheji
 ang University)
END:VEVENT
END:VCALENDAR