BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030346Z
LOCATION:Meeting Room S423+S424\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251216T131000
DTEND;TZID=Asia/Hong_Kong:20251216T141500
UID:siggraphasia_SIGGRAPH Asia 2025_sess121@linklings.com
SUMMARY:Image Restoration, Editing & Enhancement
DESCRIPTION:The Technical Papers program is the heartbeat of SIGGRAPH Asia
 , spotlighting world-class scholarly research at the forefront of computer
  graphics and interactive techniques. For decades, it has been the definit
 ive venue where bold ideas take root, foundational concepts are reimagined
 , and the future of visual computing is shaped.\n\nThis year, we explore n
 ew intersections of algorithms and artistry, automation and authorship, to
 ols and imagination – challenging the very way we design, simulate, visual
 ize, and interact with digital worlds.\n\nHRC-Net: Learning Visual Hypothe
 sis, Representative, and Collaboration for Multi-Domain Image Inpainting\n
 \nMulti-domain image inpainting utilizes complementary contextual informat
 ion from auxilliary domain images to restore corrupted regions. While exis
 ting methods reconstruct auxiliary images to provide additional guidance, 
 they face fundamental limitations: recovered pixels with complex patterns 
 often ...\n\n\nXin Wang (The Hong Kong Polytechnic University), Di Lin (Ti
 anjin University), Wanchao Su (Monash University), Ji Du and Renjie Zhang 
 (The Hong Kong Polytechnic University), Jie Zhang (Macao Polytechnic Unive
 rsity), Haotian Dong (Tianjin University), Ke Xu (City University of Hong 
 Kong), Qing Guo (Nankai University), and Ping Li (The Hong Kong Polytechni
 c University)\n---------------------\nSelf-supervised Texture Filtering\n\
 nDecomposing an image Iinto the combination of structure S and texture T c
 omponents is an important problem in computational photography and image a
 nalysis. Traditional solutions are basically non-learning based, because i
 t is difficult to construct datasets containing ground-truth decomposition
 s or ...\n\n\nHao Jiang and Rongjia Zheng (Sun Yat-sen University), Yongwe
 i Nie (South China University of Technology), Chunxia Xiao (Wuhan Universi
 ty), and Qing Zhang (Sun Yat-sen University)\n---------------------\nSelf-
 supervised Underwater Color Restoration via Wavelet-Diffusion Model with F
 iltered Multi-Scale Feature Distillation\n\nExisting underwater image proc
 essing methods often struggle due to the limited availability of real pair
 ed training data. Models trained on public datasets frequently fail to gen
 eralize across diverse underwater conditions and produce suboptimal color 
 restoration. To address these challenges, we pro...\n\n\nXin Zhang and zhu
 ang Zhou (Beijing Institute of Technology; Beijing Institute of Technology
 , Zhuhai); Yixiao Yang (National University of Singapore); Haijun Xie (Bei
 jing Institute of Technology, Zhuhai); and Haowen Yan, Hexiang Zhai, and B
 inghua Su (Beijing Institute of Technology; Beijing Institute of Technolog
 y, Zhuhai)\n---------------------\nELAD: Blind Face Restoration using Expe
 ctation-based Likelihood Approximation and Diffusion Prior\n\nBlind Face R
 estoration (BFR) aims to recover face images suffering from unknown degrad
 ations. A recent approach to solve BFR is via plug-and-play methods for im
 age restoration, which combine a likelihood function with pre-trained diff
 usion models as priors. However, as the likelihood is inherently u...\n\n\
 nSean Man, Guy Ohayon, Ron Raphaeli, and Matan Kleiner (Technion – Israel 
 Institute of Technology) and Michael Elad (Technion – Israel Institute of 
 Technology, NVIDIA)\n---------------------\nDvD: Unleashing a Generative P
 aradigm for Document Dewarping via Coordinates-based Diffusion Model\n\nDo
 cument dewarping aims to rectify deformations in photographic document ima
 ges, thus improving text readability, which has attracted much attention a
 nd made great progress, but it is still challenging to preserve document s
 tructures. Given recent advances in diffusion models, it is natural for us
  t...\n\n\nWeiguang Zhang, Huangcheng Lu, and Maizhen Ning (Xi’an Jiaotong
 -Liverpool University, University of Liverpool); Xiaowei Huang (University
  of Liverpool); Wei Wang (Xi’an Jiaotong-Liverpool University); Kaizhu Hua
 ng (Duke Kunshan University); and Qiufeng Wang (Xi’an Jiaotong-Liverpool U
 niversity)\n---------------------\nHarnessing Diffusion-Yielded Score Prio
 rs for Image Restoration\n\nDeep image restoration models aim to learn a m
 apping from degraded image space to natural image space. However, they fac
 e several critical challenges: removing degradation, generating realistic 
 details, and ensuring pixel-level consistency. Over time, three major clas
 ses of methods have emerged, inc...\n\n\nXinqi Lin (Shenzhen Institute of 
 Advanced Technology，Chinese Academy of Sciences, University of Chinese Aca
 demy of Sciences); Fanghua Yu (Shenzhen Institute of Advanced Technology，C
 hinese Academy of Sciences); Jinfan Hu (Shenzhen Institutes of Advanced Te
 chnology，Chinese Academy of Sciences, University of Chinese Academy of Sci
 ences); Zhiyuan You (Shenzhen Institutes of Advanced Technology，Chinese Ac
 ademy of Sciences, The Chinese University of Hong Kong); Wu Shi (Shenzhen 
 Institutes of Advanced Technology，Chinese Academy of Sciences); Jimmy S. R
 en (Hong Kong Metropolitan University, SenseTime Research); Jinjin Gu (INS
 AIT, Sofia University); and Chao Dong (Shenzhen Institutes of Advanced Tec
 hnology，Chinese Academy of Sciences)\n\nRegistration Category: Full Access
 , Full Access Supporter\n\nSession Chair: Paul Debevec (Eyeline, USC Insti
 tute for Creative Technologies (ICT))
END:VEVENT
END:VCALENDAR