BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260114T163717Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231214T163500
DTEND;TZID=Australia/Melbourne:20231214T164500
UID:siggraphasia_SIGGRAPH Asia 2023_sess133_papers_538@linklings.com
SUMMARY:MatFusion: A Generative Diffusion Model for SVBRDF Capture
DESCRIPTION:Sam Sartor and Pieter Peers (College of William & Mary)\n\nWe 
 formulate SVBRDF estimation from photographs as a diffusion task. To model
  the distribution of spatially varying materials, we first train a novel u
 nconditional SVBRDF diffusion backbone model on a large set of 312,165 syn
 thetic spatially varying material exemplars.  This SVBRDF  diffusion backb
 one model, named MatFusion, can then serve as a basis for refining a condi
 tional diffusion model to estimate the material properties from a photogra
 ph under controlled or uncontrolled lighting. Our backbone MatFusion model
  is trained using only a loss on the reflectance properties,  and therefor
 e refinement can be paired with more expensive rendering methods without t
 he need for backpropagation during training.  Because the conditional SVBR
 DF diffusion models are generative, we can synthesize  multiple SVBRDF est
 imates from the same input photograph from which the user can select the o
 ne that best matches the users' expectation.  We demonstrate the flexibili
 ty of our method by refining different SVBRDF diffusion models conditioned
  on different types of incident lighting, and show that for a single photo
 graph under colocated flash lighting our method achieves equal or better a
 ccuracy than existing SVBRDF estimation methods.\n\nRegistration Category:
  Full Access\n\nSession Chair: Anton Kaplanyan (Intel)\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_538&sess=sess133
END:VEVENT
END:VCALENDAR