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:20240214T070250Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231215T151500
DTEND;TZID=Australia/Melbourne:20231215T152500
UID:siggraphasia_SIGGRAPH Asia 2023_sess138_papers_433@linklings.com
SUMMARY:UVDoc: Neural Grid-based Document Unwarping
DESCRIPTION:Technical Papers\n\nFloor Verhoeven, Tanguy Magne, and Olga So
 rkine-Hornung (ETH Zurich)\n\nRestoring the original, flat appearance of a
  printed document from casual photographs of bent and wrinkled pages is a 
 common everyday problem. In this paper we propose a novel method for grid-
 based single-image document unwarping. Our method performs geometric disto
 rtion correction via a fully convolutional deep neural network that learns
  to predict the 3D grid mesh of the document and the corresponding 2D unwa
 rping grid in a multi-task fashion, implicitly encoding the coupling betwe
 en the shape of a 3D piece of paper and its 2D image. In order to allow un
 warping models to train on data that is more realistic in appearance than 
 the commonly used synthetic Doc3D dataset we create and publish our own da
 taset, called UVDoc, which combines pseudo-photorealistic document images 
 with physically accurate 3D shape and unwarping function annotations. Our 
 dataset is labeled with all the information necessary to train our unwarpi
 ng network, without having to engineer separate loss functions that can de
 al with the lack of ground-truth typically found in document in the wild d
 atasets. We perform an in-depth evaluation that demonstrates that with the
  inclusion of our novel pseudo-photorealistic dataset, our relatively smal
 l network architecture achieves state-of-the-art results on the DocUNet be
 nchmark. We show that the pseudo-photorealistic nature of our UVDoc datase
 t allows for new and better evaluation methods, such as lighting-corrected
  MS-SSIM. We provide a novel benchmark dataset that facilitates such evalu
 ations, and propose a metric that quantifies line straightness after unwar
 ping. Our code, results and UVDoc dataset will be made publicly available 
 upon publication.\n\nRegistration Category: Full Access\n\nSession Chair: 
 Seung-Hwan Baek (POSTECH)
URL:https://asia.siggraph.org/2023/full-program?id=papers_433&sess=sess138
END:VEVENT
END:VCALENDAR