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:20230103T035312Z
LOCATION:Room 324\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221208T170000
DTEND;TZID=Asia/Seoul:20221208T183000
UID:siggraphasia_SIGGRAPH Asia 2022_sess167_papers_247@linklings.com
SUMMARY:Metappearance: Meta-Learning for Visual Appearance Reproduction
DESCRIPTION:Technical Communications, Technical Papers\n\nMetappearance: M
 eta-Learning for Visual Appearance Reproduction\n\nFischer, Ritschel\n\nTh
 ere currently exist two main approaches to reproducing visual appearance u
 sing Machine Learning (ML): The first is training models that generalize o
 ver different instances of a problem, e.g., different images of a dataset.
  As one-shot approaches, these offer fast inference, but often fall short 
 in quality. The second approach does not train models that generalize acro
 ss tasks, but rather over-fit a single instance of a problem, e.g., a flas
 h image of a material. These methods offer high quality, but take long to 
 train. We suggest to combine both techniques end-to-end using meta-learnin
 g: We over-fit onto a single problem instance in an inner loop, while also
  learning how to do so efficiently in an outer-loop across many exemplars.
  To this end, we derive the required formalism that allows applying meta-l
 earning to a wide range of visual appearance reproduction problems: textur
 es, Bi-directional Reflectance Distribution Functions (BRDFs), spatially-v
 arying\nBRDFs (svBRDFs), illumination or the entire light transport of a s
 cene. The effects of meta-learning parameters on several different aspects
  of visual appearance are analyzed in our framework, and specific guidance
  for different tasks is provided. Metapperance enables visual quality that
  is similar to over-fit approaches in only a fraction of their runtime whi
 le keeping the adaptivity of general models.\n\nRegistration Category: FUL
 L ACCESS, ON-DEMAND ACCESS\n\nLanguage: ENGLISH\n\nFormat: IN-PERSON, ON-D
 EMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_247&sess=sess16
 7
END:VEVENT
END:VCALENDAR