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:20251218T030653Z
LOCATION:Meeting Room S423+S424\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251218T110100
DTEND;TZID=Asia/Hong_Kong:20251218T111200
UID:siggraphasia_SIGGRAPH Asia 2025_sess151_papers_1525@linklings.com
SUMMARY:Virtually Being: Customizing Camera-Controllable Video Diffusion M
 odels with Volumetric Performance Captures
DESCRIPTION:Yuancheng Xu, Wenqi Xian, Li Ma, Julien Philip, and Ahmet Taşe
 l (Eyeline); Yiwei Zhao (Netflix); Ryan Burgert (Eyeline, Stony Brook Univ
 ersity); Mingming He, Oliver Hermann, and Oliver Pilarski (Eyeline); Rahul
  Garg (Netflix); and Paul Debevec and Ning Yu (Eyeline)\n\nWe introduce a 
 framework that enables both multi-view character consistency and 3D camera
  control in video diffusion models through a novel customization data pipe
 line.  We train the character consistency component with recorded volumetr
 ic capture performances re-rendered with diverse camera trajectories via 4
 D Gaussian Splatting (4DGS), lighting variability obtained with a video re
 lighting model.  We fine-tune state-of-the-art open-source video diffusion
  models on this data to provide strong identity preservation, precise came
 ra control, and lighting adaptability.  Our framework also supports core c
 apabilities for virtual production, including multi-subject generation usi
 ng two approaches: joint training and noise blending, the latter enabling 
 efficient composition of independently customized models at inference time
 ; \nit also achieves scene and real-life video customization as well as co
 ntrol over motion and spatial layout during customization. Extensive exper
 iments show improved video quality, higher personalization accuracy, and e
 nhanced camera control and lighting adaptability, advancing the integratio
 n of AI-driven video generation into virtual production.\n\nRegistration C
 ategory: Full Access, Full Access Supporter\n\nSession Chair: Min Lu (Shen
 zhen University (SZU))\n\n
END:VEVENT
END:VCALENDAR