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:20240214T070248Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231214T165300
DTEND;TZID=Australia/Melbourne:20231214T170300
UID:siggraphasia_SIGGRAPH Asia 2023_sess133_papers_812@linklings.com
SUMMARY:Diffusion-based Holistic Texture Rectification and Synthesis
DESCRIPTION:Technical Communications, Technical Papers\n\nGuoqing Hao (Uni
 versity of Tsukuba, National Institute of Advanced Industrial Science and 
 Technology); Satoshi Iizuka (University of Tsukuba); Kensho Hara (National
  Institute of Advanced Industrial Science); Edgar Simo-Serra (Waseda Unive
 rsity); Hirokatsu Kataoka (National Institute of Advanced Industrial Scien
 ce); and Kazuhiro Fukui (University of Tsukuba)\n\nWe present a novel fram
 ework for rectifying occlusions and distortions in degraded texture sample
 s from natural images. Traditional texture synthesis approaches focus on g
 enerating textures from pristine samples, which necessitate meticulous pre
 paration by humans and are often unattainable in most natural images. Thes
 e challenges stem from the frequent occlusions and distortions of texture 
 samples in natural images due to obstructions and variations in object sur
 face geometry. To address these issues, we propose a framework that synthe
 sizes holistic textures from degraded samples in natural images, extending
  the applicability of exemplar-based texture synthesis techniques. Our fra
 mework utilizes a conditional Latent Diffusion Model (LDM) with a novel oc
 clusion-aware latent transformer. This latent transformer not only effecti
 vely encodes texture features from partially-observed samples necessary fo
 r the generation process of the LDM, but also explicitly captures long-ran
 ge dependencies in samples with large occlusions. To train our model, we i
 ntroduce a method for generating synthetic data by applying geometric tran
 sformations and free-form mask generation to clean textures. Experimental 
 results demonstrate that our framework significantly outperforms existing 
 methods both quantitatively and quantitatively. Furthermore, we conduct co
 mprehensive ablation studies to validate the different components of our p
 roposed framework. Results are corroborated by a perceptual user study whi
 ch highlights the efficiency of our proposed approach.\n\nRegistration Cat
 egory: Full Access\n\nSession Chair: Anton Kaplanyan (Intel)
URL:https://asia.siggraph.org/2023/full-program?id=papers_812&sess=sess133
END:VEVENT
END:VCALENDAR