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:20250110T023313Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241206T105600
DTEND;TZID=Asia/Tokyo:20241206T110800
UID:siggraphasia_SIGGRAPH Asia 2024_sess143_papers_468@linklings.com
SUMMARY:GaussianHeads: End-to-End Learning of Drivable Gaussian Head Avata
 rs from Coarse-to-fine Representations
DESCRIPTION:Technical Papers\n\nKartik Teotia (Max Planck Institute for In
 formatics, Saarland Informatics Campus); Hyeongwoo Kim (Imperial College L
 ondon); Pablo Garrido (Flawless AI); Marc Habermann (Max Planck Institute 
 for Informatics, Saarland Informatics Campus); Mohamed Elgharib (Max Planc
 k Institute for Informatics); and Christian Theobalt (Max Planck Institute
  for Informatics, Saarland Informatics Campus)\n\nReal-time rendering of h
 uman head avatars is a cornerstone of many computer graphics applications,
  such as augmented reality, video games, and films, to name a few. Recent 
 approaches address this challenge with computationally efficient geometry 
 primitives in a carefully calibrated multi-view setup. Albeit producing ph
 otorealistic head renderings, it often fails to represent complex motion c
 hanges such as the mouth interior and strongly varying head poses. We prop
 ose a new method to generate highly dynamic and deformable human head avat
 ars from multi-view imagery in real-time. At the core of our method is a h
 ierarchical representation of head models that allows to capture the compl
 ex dynamics of facial expressions and head movements. First, with rich fac
 ial features extracted from raw input frames, we learn to deform the coars
 e facial geometry of the template mesh. We then initialize 3D Gaussians on
  the deformed surface and refine their positions in a fine step. We train 
 this coarse-to-fine facial avatar model along with the head pose as a lear
 nable parameter in an end-to-end framework. This enables not only controll
 able facial animation via video inputs, but also high-fidelity novel view 
 synthesis of challenging facial expressions, such as tongue deformations a
 nd fine-grained teeth structure under large motion changes. Moreover, it e
 ncourages the learned head avatar to generalize towards new facial express
 ions and head poses at inference time. We demonstrate the performance of o
 ur method with comparisons against the related methods on different datase
 ts, spanning challenging facial expression sequences across multiple ident
 ities. We also show the potential application of our approach by demonstra
 ting a cross-identity facial performance transfer application.\n\nRegistra
 tion Category: Full Access, Full Access Supporter\n\nLanguage Format: Engl
 ish Language\n\nSession Chair: Iain Matthews (Epic Games, Carnegie Mellon 
 University)
URL:https://asia.siggraph.org/2024/program/?id=papers_468&sess=sess143
END:VEVENT
END:VCALENDAR