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:20230103T035307Z
LOCATION:Auditorium\, Level 5\, West Wing
DTSTART;TZID=Asia/Seoul:20221206T100000
DTEND;TZID=Asia/Seoul:20221206T120000
UID:siggraphasia_SIGGRAPH Asia 2022_sess153_papers_469@linklings.com
SUMMARY:Rapid Face Asset Acquisition with Recurrent Feature Alignment
DESCRIPTION:Technical Papers\n\nRapid Face Asset Acquisition with Recurren
 t Feature Alignment\n\nLiu, Cai, Chen, Zhou, Zhao\n\nWe present Recurrent 
 Feature Alignment (ReFA), an end-to-end neural network for the very rapid 
 creation of production-grade face assets from multi-view images. ReFA is o
 n par with the industrial pipelines in quality for producing accurate, com
 plete, registered, and textured assets directly applicable to physically-b
 ased rendering, but produces the asset end-to-end, fully automatically at 
 a significantly faster speed at 4.5 FPS, which is un-\nprecedented among n
 eural-based techniques. Our method represents face 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 iter
 atively optimizes the geometry by projecting the image-space features to t
 he UV space and comparing them with a reference UV-space feature. The opti
 mized geometry then provides pixel-aligned signals for the inference of hi
 gh-resolution textures. Experiments have validated that ReFA achieves a me
 dian error of 0.603𝑚𝑚 in geometry reconstruction, is robust to extreme pos
 e and expression, and excels in sparse-view settings. We believe that the 
 progress achieved by our network enables lightweight, fast face assets acq
 uisition that significantly boosts the downstream applications, such as av
 atar creation and facial performance capture. It will also enable massive 
 database capturing for deep learning purposes.\n\nRegistration Category: F
 ULL ACCESS, EXPERIENCE PLUS ACCESS, EXPERIENCE ACCESS, TRADE EXHIBITOR\n\n
 Language: ENGLISH\n\nFormat: IN-PERSON
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_469&sess=sess15
 3
END:VEVENT
END:VCALENDAR