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:20250110T023312Z
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:Technical Papers\n\nAditya Ganeshan, Ryan Huang, Xianghao Xu, 
 R. Kenny Jones, and Daniel Ritchie (Brown University)\n\nThe ability to ed
 it 3D assets from natural language presents a compelling paradigm to aid i
 n the democratization of 3D content creation. However, while natural langu
 age is often effective at communicating general intent, it is poorly suite
 d for specifying exact manipulation. To address this gap, we introduce Par
 SEL, a system that enables controllable editing of high-quality 3D assets 
 from natural language. Given a segmented 3D mesh and an editing request, P
 arSEL produces a parameterized editing program. Adjusting these parameters
  allows users to explore shape variations with exact control over the magn
 itudes of edits. To infer editing programs which align with an input edit 
 request, we leverage the abilities of large-language models (LLMs). Howeve
 r, while we find that LLMs excel at identifying initial edit operations, t
 hey often fail to infer complete editing programs, and produce outputs tha
 t violate shape semantics. To overcome this issue, we introduce Analytical
  Edit Propagation (AEP), an algorithm which extends a seed edit with addit
 ional operations until a complete editing program has been formed. Unlike 
 prior methods, AEP searches for analytical editing operations compatible w
 ith a range of possible user edits through the integration of computer alg
 ebra systems for geometric analysis. Experimentally we demonstrate ParSEL'
 s effectiveness in enabling controllable editing of 3D objects through nat
 ural language requests over alternative system designs.\n\nRegistration Ca
 tegory: Full Access, Full Access Supporter\n\nLanguage Format: English Lan
 guage\n\nSession Chair: Kai Wang (Amazon)
URL:https://asia.siggraph.org/2024/program/?id=papers_441&sess=sess114
END:VEVENT
END:VCALENDAR
