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:20230103T035309Z
LOCATION:Room 325-AB\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221207T153000
DTEND;TZID=Asia/Seoul:20221207T170000
UID:siggraphasia_SIGGRAPH Asia 2022_sess163_papers_469@linklings.com
SUMMARY:Rapid Face Asset Acquisition with Recurrent Feature Alignment
DESCRIPTION:Technical Communications, Technical Papers\n\nRapid Face Asset
  Acquisition with Recurrent Feature Alignment\n\nLiu, Cai, Chen, Zhou, Zha
 o\n\nWe present Recurrent Feature Alignment (ReFA), an end-to-end neural n
 etwork for the very rapid creation of production-grade face assets from mu
 lti-view images. ReFA is on par with the industrial pipelines in quality f
 or producing accurate, complete, registered, and textured assets directly 
 applicable to physically-based rendering, but produces the asset end-to-en
 d, fully automatically at a significantly faster speed at 4.5 FPS, which i
 s un-\nprecedented among neural-based techniques. Our method represents fa
 ce geometry as a position map in the UV space. The network first extracts 
 per-pixel features in both the multi-view image space and the UV space. A 
 recurrent module then iteratively optimizes the geometry by projecting the
  image-space features to the UV space and comparing them with a reference 
 UV-space feature. The optimized geometry then provides pixel-aligned signa
 ls for the inference of high-resolution textures. Experiments have validat
 ed that ReFA achieves a median error of 0.603𝑚𝑚 in geometry reconstruction
 , is robust to extreme pose and expression, and excels in sparse-view sett
 ings. We believe that the progress achieved by our network enables lightwe
 ight, fast face assets acquisition that significantly boosts the downstrea
 m applications, such as avatar creation and facial performance capture. It
  will also enable massive database capturing for deep learning purposes.\n
 \nRegistration Category: FULL ACCESS, ON-DEMAND ACCESS\n\nLanguage: ENGLIS
 H\n\nFormat: IN-PERSON, ON-DEMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_469&sess=sess16
 3
END:VEVENT
END:VCALENDAR