BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260817T171530Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241204T170500
DTEND;TZID=Asia/Tokyo:20241204T171600
UID:siggraphasia_SIGGRAPH Asia 2024_sess122_papers_369@linklings.com
SUMMARY:TEXGen: a Generative Diffusion Model for Mesh Textures
DESCRIPTION:Xin Yu (University of Hong Kong); Ze Yuan (Beihang University)
 ; Yuan-Chen Guo (VAST); Ying-Tian Liu (Tsinghua University); Jianhui Liu (
 University of Hong Kong); Yangguang Li, Yan-Pei Cao, and Ding Liang (VAST)
 ; and Xiaojuan Qi (University of Hong Kong)\n\nWhile high-quality texture 
 maps are essential for realistic 3D asset rendering, few studies have expl
 ored learning directly in the texture space, especially on large-scale dat
 asets. In this work, we depart from the conventional approach of relying o
 n pre-trained 2D diffusion models for test-time optimization of 3D texture
 s. Instead, we focus on the fundamental problem of learning in the UV text
 ure space itself. For the first time, we train a large diffusion model cap
 able of directly generating high-resolution texture maps in a feed-forward
  manner.\nTo facilitate efficient learning in high-resolution UV spaces, w
 e propose a scalable network architecture that interleaves convolutions on
  UV maps with attention layers on point clouds. Leveraging this architectu
 ral design, we train a 700 million parameter diffusion model that can gene
 rate UV texture maps guided by text prompts and single-view images. Once t
 rained, our model naturally supports various extended applications, includ
 ing text-guided texture inpainting, sparse-view texture completion, and te
 xt-driven texture synthesis.\n\nRegistration Category: Full Access, Full A
 ccess Supporter\n\nLanguage Format: English Language\n\nSession Chair: Min
 hyuk Sung (Korea Advanced Institute of Science and Technology (KAIST))\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_369&sess=sess122
END:VEVENT
END:VCALENDAR