BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Seoul
X-LIC-LOCATION:Asia/Seoul
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:KST
DTSTART:18871231T000000
DTSTART:19881009T020000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230103T035308Z
LOCATION:Room 325-AB\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221207T110000
DTEND;TZID=Asia/Seoul:20221207T123000
UID:siggraphasia_SIGGRAPH Asia 2022_sess161_papers_398@linklings.com
SUMMARY:FDNeRF: Few-shot Dynamic Neural Radiance Fields for Face Reconstru
 ction and Expression Editing
DESCRIPTION:Technical Communications, Technical Papers\n\nFDNeRF: Few-shot
  Dynamic Neural Radiance Fields for Face Reconstruction and Expression Edi
 ting\n\nZHANG, LI, WAN, WANG, LIAO\n\nWe propose a Few-shot Dynamic Neural
  Radiance Field (FDNeRF), the first NeRF-based method capable of reconstru
 ction and expression editing of 3D faces based on a small number of dynami
 c images. Unlike existing dynamic NeRFs that require dense images as input
  and can only be modeled for a single identity, our method enables face re
 construction across different persons with few-shot inputs. Compared to st
 ate-of-the-art few-shot NeRFs designed for modeling static scenes, the pro
 posed FDNeRF accepts view-inconsistent dynamic inputs and supports arbitra
 ry facial expression editing, i.e., producing faces with novel expressions
  beyond the input ones. To handle the inconsistencies between dynamic inpu
 ts, we introduce a well-designed conditional feature warping (CFW) module 
 to perform expression conditioned warping in 2D feature space, which is al
 so identity adaptive and 3D constrained. As a result, features of differen
 t expressions are transformed into the target ones. We then construct a ra
 diance field based on these view-consistent features and use volumetric re
 ndering to synthesize novel views of the modeled faces. Extensive experime
 nts with quantitative and qualitative evaluation demonstrate that our meth
 od outperforms existing dynamic and few-shot NeRFs on both 3D face reconst
 ruction and expression editing tasks. Code is available at https://github.
 com/FDNeRF/FDNeRF.\n\nRegistration Category: FULL ACCESS, ON-DEMAND ACCESS
 \n\nLanguage: ENGLISH\n\nFormat: IN-PERSON, ON-DEMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_398&sess=sess16
 1
END:VEVENT
END:VCALENDAR