BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030657Z
LOCATION:Meeting Room S421\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251218T092100
DTEND;TZID=Asia/Hong_Kong:20251218T093200
UID:siggraphasia_SIGGRAPH Asia 2025_sess146_papers_1208@linklings.com
SUMMARY:LARM: A Large Articulated Object Reconstruction Model
DESCRIPTION:Sylvia Yuan, Ruoxi Shi, and Xinyue Wei (University of Californ
 ia San Diego); Xiaoshuai Zhang (Hillbot); Hao Su (University of California
  San Diego); and Minghua Liu (Hillbot)\n\nModeling 3D articulated objects 
 with realistic geometry, textures, and kinematics is essential for a wide 
 range of applications. However, existing optimization-based reconstruction
  methods often require dense multi-view inputs and expensive per-instance 
 optimization, limiting their scalability. Recent feedforward approaches of
 fer faster alternatives but frequently produce coarse geometry, lack textu
 re reconstruction, and rely on brittle, complex multi-stage pipelines. We 
 introduce LARM, a unified feedforward framework that reconstructs 3D artic
 ulated objects from sparse-view images by jointly recovering detailed geom
 etry, realistic textures, and accurate joint structures. LARM extends LVSM
 —a recent novel view synthesis (NVS) approach for static 3D objects—into t
 he articulated setting by jointly reasoning over camera pose and articulat
 ion variation using a transformer-based architecture, enabling scalable an
 d accurate novel view synthesis. In addition, LARM generates auxiliary out
 puts such as depth maps and part masks to facilitate explicit 3D mesh extr
 action and joint estimation. Our pipeline eliminates the need for dense su
 pervision and supports high-fidelity reconstruction across diverse object 
 categories. Extensive experiments demonstrate that LARM outperforms state-
 of-the-art methods in both novel view and state synthesis as well as 3D ar
 ticulated object reconstruction, generating high-quality meshes that close
 ly adhere to the input images.\n\nRegistration Category: Full Access, Full
  Access Supporter\n\nSession Chair: Ziqi Wang (HKUST, EPFL)\n\n
END:VEVENT
END:VCALENDAR