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_383@linklings.com
SUMMARY:Look-Ahead Training with Learned Reflectance Loss for Single-Image
  SVBRDF Estimation
DESCRIPTION:Technical Papers\n\nLook-Ahead Training with Learned Reflectan
 ce Loss for Single-Image SVBRDF Estimation\n\nZhou, Kalantari\n\nIn this p
 aper, we propose a novel optimization-based method to estimate the reflect
 ance properties of a near planar surface from a single input image. Specif
 ically, we perform test-time optimization by directly updating the paramet
 ers of a neural network to minimize the test error. Since single image SVB
 RDF estimation is highly ill-posed, such an optimization is prone to\nover
 fitting. Our main contribution is to address this problem by introducing a
  training mechanism that takes the test-time optimization into account. Sp
 ecifically, we train our network by minimizing the training loss after one
  or more gradient updates with the test loss. By training the network in t
 his manner, we ensure that the network does not overfit to the input image
  during the test-time optimization process. Additionally, we propose a lea
 rned reflectance loss to augment the typically used rendering loss during 
 the test-time optimization. We do so by using an auxiliary network that es
 timates pseudo ground truth reflectance parameters and train it in combina
 tion with the main network. Our approach is able to converge with a small 
 number of iterations of the test-time optimization and produces better res
 ults compared to the state-of-the-art methods.\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_383&sess=sess15
 3
END:VEVENT
END:VCALENDAR