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:20260817T171532Z
LOCATION:Hall B7 (1)\, B Block\, Level 7
DTSTART;TZID=Asia/Tokyo:20241204T111900
DTEND;TZID=Asia/Tokyo:20241204T113100
UID:siggraphasia_SIGGRAPH Asia 2024_sess114_papers_441@linklings.com
SUMMARY:ParSEL: Parameterized Shape Editing with Language
DESCRIPTION:Aditya Ganeshan, Ryan Huang, Xianghao Xu, R. Kenny Jones, and 
 Daniel Ritchie (Brown University)\n\nThe ability to edit 3D assets from na
 tural language presents a compelling paradigm to aid in the democratizatio
 n of 3D content creation. However, while natural language is often effecti
 ve at communicating general intent, it is poorly suited for specifying exa
 ct manipulation. To address this gap, we introduce ParSEL, a system that e
 nables controllable editing of high-quality 3D assets from natural languag
 e. Given a segmented 3D mesh and an editing request, ParSEL produces a par
 ameterized editing program. Adjusting these parameters allows users to exp
 lore shape variations with exact control over the magnitudes of edits. To 
 infer editing programs which align with an input edit request, we leverage
  the abilities of large-language models (LLMs). However, while we find tha
 t LLMs excel at identifying initial edit operations, they often fail to in
 fer complete editing programs, and produce outputs that violate shape sema
 ntics. To overcome this issue, we introduce Analytical Edit Propagation (A
 EP), an algorithm which extends a seed edit with additional operations unt
 il a complete editing program has been formed. Unlike prior methods, AEP s
 earches for analytical editing operations compatible with a range of possi
 ble user edits through the integration of computer algebra systems for geo
 metric analysis. Experimentally we demonstrate ParSEL's effectiveness in e
 nabling controllable editing of 3D objects through natural language reques
 ts over alternative system designs.\n\nRegistration Category: Full Access,
  Full Access Supporter\n\nLanguage Format: English Language\n\nSession Cha
 ir: Kai Wang (Simon Fraser University, ShanghaiTech University)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_441&sess=sess114
END:VEVENT
END:VCALENDAR