BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260114T163651Z
LOCATION:Meeting Room C4.8\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231214T092500
DTEND;TZID=Australia/Melbourne:20231214T094000
UID:siggraphasia_SIGGRAPH Asia 2023_sess148_papers_257@linklings.com
SUMMARY:TwinTex: Geometry-aware Texture Generation for Abstracted 3D Archi
 tectural Models
DESCRIPTION:Weidan Xiong, Hongqian Zhang, Botao Peng, Ziyu Hu, Yongli Wu, 
 Jianwei Guo, and Hui Huang (Shenzhen University)\n\nCoarse architectural m
 odels are often generated at scales ranging from individual buildings to s
 cenes for downstream applications such as Digital Twin City, Metaverse, LO
 Ds, etc. Such piece-wise planar models can be abstracted as twins from 3D 
 dense reconstructions. However, these models typically lack realistic text
 ure relative to the real building or scene, making them unsuitable for viv
 id display or direct reference. In this paper, we present TwinTex, the fir
 st automatic texture mapping framework to generate a photo-realistic textu
 re for a piece-wise planar proxy. Our method addresses most challenges occ
 urring in such twin texture generation. Specifically, for each primitive p
 lane, we first select a small set of photos with greedy heuristics conside
 ring photometric quality, perspective quality and facade texture completen
 ess. Then, different levels of line features (LoLs) are extracted from the
  set of selected photos to generate guidance for later steps. With LoLs, w
 e employ optimization algorithms to align texture with geometry from local
  to global. Finally, we fine-tune a diffusion model with a multi-mask init
 ialization component and a new dataset to inpaint the missing region. Expe
 rimental results on many buildings, indoor scenes and man-made objects of 
 varying complexity demonstrate the generalization ability of our algorithm
 . Our approach surpasses state-of-the-art texture mapping methods in terms
  of high-fidelity quality and reaches a human-expert production level with
  much less effort.\n\nRegistration Category: Full Access\n\nSession Chair:
  Lin Lu (Shandong University)\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_257&sess=sess148
END:VEVENT
END:VCALENDAR