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:20260817T171534Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241206T111900
DTEND;TZID=Asia/Tokyo:20241206T113100
UID:siggraphasia_SIGGRAPH Asia 2024_sess143_papers_454@linklings.com
SUMMARY:Robust Dual Gaussian Splatting for Immersive Human-centric Volumet
 ric Videos
DESCRIPTION:Yuheng Jiang, Zhehao Shen, Yu Hong, Chengcheng Guo, and Yize W
 u (ShanghaiTech University); Yingliang Zhang (DGene Inc.); and Jingyi Yu a
 nd Lan Xu (ShanghaiTech University)\n\nVolumetric video represents a trans
 formative advancement in visual media, enabling users to freely navigate i
 mmersive virtual experiences and narrowing the gap between digital and rea
 l worlds. However, the need for extensive manual intervention to stabilize
  mesh sequences and the generation of excessively large assets in existing
  workflows impedes broader adoption.\nIn this paper, we present a novel Ga
 ussian-based approach, dubbed DualGS, for real-time and high-fidelity play
 back of complex human performance with excellent compression ratios. Our k
 ey idea in DualGS is to separately represent motion and appearance using t
 he corresponding skin and joint Gaussians. Such an explicit disentanglemen
 t can significantly reduce motion redundancy and enhance temporal coherenc
 e. We begin by initializing the DualGS and anchoring skin Gaussians to joi
 nt Gaussians at the first frame. Subsequently, we employ a coarse-to-fine 
 training strategy for frame-by-frame human performance modeling. It includ
 es a coarse alignment phase for overall motion prediction as well as a fin
 e-grained optimization for robust tracking and high-fidelity rendering. To
  integrate volumetric video seamlessly into VR environments, we efficientl
 y compress motion using entropy encoding and appearance using codec compre
 ssion coupled with a persistent codebook. Our approach achieves a compress
 ion ratio of up to 120 times, only requiring approximately 350KB of storag
 e per frame. We demonstrate the efficacy of our representation through pho
 to-realistic, free-view experiences on VR headsets, enabling users to imme
 rsively watch musicians in performance and feel the rhythm of the notes at
  the performers' fingertips.\n\nRegistration Category: Full Access, Full A
 ccess Supporter\n\nLanguage Format: English Language\n\nSession Chair: Iai
 n Matthews (Epic Games, Carnegie Mellon University)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_454&sess=sess143
END:VEVENT
END:VCALENDAR