BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240214T070240Z
LOCATION:Darling Harbour Theatre\, Level 2 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T093000
DTEND;TZID=Australia/Melbourne:20231212T124500
UID:siggraphasia_SIGGRAPH Asia 2023_sess209_papers_891@linklings.com
SUMMARY:BakedAvatar: Baking Neural Fields for Real-Time Head Avatar Synthe
 sis
DESCRIPTION:Technical Papers\n\nHao-Bin Duan (Beihang University); Miao Wa
 ng (Beihang University, Zhongguancun Laboratory); Jin-Chuan Shi and Xu-Chu
 an Chen (Beihang University); and Yan-Pei Cao (Tencent)\n\nSynthesizing ph
 otorealistic 4D human head avatars from videos is essential for VR/AR, tel
 epresence, and video game applications. Although existing Neural Radiance 
 Fields (NeRF)-based methods achieve high-fidelity results, the computation
 al expense limits their use in real-time applications. To overcome this li
 mitation, we introduce BakedAvatar, a novel representation for real-time n
 eural head avatar synthesis, deployable in a standard polygon rasterizatio
 n pipeline. Our approach extracts deformable multi-layer meshes from learn
 ed isosurfaces of the head and computes expression-, pose-, and view-depen
 dent appearances that can be baked into static textures for efficient rast
 erization. We thus propose a three-stage pipeline for neural head avatar s
 ynthesis, which includes learning continuous deformation, manifold, and ra
 diance fields, extracting layered meshes and textures, and fine-tuning tex
 ture details with differential rasterization. Experimental results demonst
 rate that our representation generates synthesis results of comparable qua
 lity to other state-of-the-art methods while significantly reducing the in
 ference time required. We further showcase various head avatar synthesis r
 esults from monocular videos, including view synthesis, face reenactment, 
 expression editing, and pose editing, all at interactive frame rates.\n\nR
 egistration Category: Full Access, Enhanced Access, Trade Exhibitor, Exper
 ience Hall Exhibitor
URL:https://asia.siggraph.org/2023/full-program?id=papers_891&sess=sess209
END:VEVENT
END:VCALENDAR