BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250110T023312Z
LOCATION:Hall B5 (1)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241204T144500
DTEND;TZID=Asia/Tokyo:20241204T145900
UID:siggraphasia_SIGGRAPH Asia 2024_sess118_papers_431@linklings.com
SUMMARY:GarVerseLOD: High-Fidelity 3D Garment Reconstruction from a Single
  In-the-Wild Image using a Dataset with Levels of Details
DESCRIPTION:Technical Papers\n\nZhongjin Luo, Haolin Liu, Chenghong Li, Wa
 nghao Du, Zirong Jin, and Wanhu Sun (Chinese University of Hong Kong, Shen
 zhen); Yinyu Nie (Huawei Technologies Ltd.); Weikai Chen (Tencent America)
 ; and Xiaoguang Han (Chinese University of Hong Kong, Shenzhen)\n\nNeural 
 implicit functions have brought impressive advances to the state-of-the-ar
 t of clothed human digitization from multiple or even single images. Howev
 er, despite the progress, current arts still have difficulty generalizing 
 to unseen images with complex cloth deformation and body poses. In this wo
 rk, we present GarVerseLOD, a new dataset and framework that paves the way
  to achieving unprecedented robustness in high-fidelity 3D garment reconst
 ruction from a single unconstrained image. Inspired by the recent success 
 of large generative models, we believe that one key to addressing the gene
 ralization challenge lies in the quantity and quality of 3D garment data. 
 Towards this end, GarVerseLOD collects 6,000 high-quality cloth models wit
 h fine-grained geometry details manually created by professional artists. 
 In addition to the scale of training data, we observe that having disentan
 gled granularities of geometry can play an important role in boosting the 
 generalization capability and inference accuracy of the learned model.	We 
 hence craft GarVerseLOD as a hierarchical dataset with levels of details (
 LOD), spanning from detail-free stylized shape to pose-blended garment wit
 h pixel-aligned details. This allows us to make this highly under-constrai
 ned problem tractable by factorizing the inference into easier tasks, each
  narrowed down with smaller searching space. To ensure GarVerseLOD can gen
 eralize well to in-the-wild images, we propose a novel labeling paradigm b
 ased on conditional diffusion models to generate extensive paired images f
 or each garment model with high photorealism. We evaluate our method on a 
 massive amount of in-the-wild images. Experimental results demonstrate tha
 t GarVerseLOD can generate standalone garment pieces with significantly be
 tter quality than prior approaches while being robust against a large vari
 ation of pose, illumination, occlusion, and deformation. Code and dataset 
 are available at garverselod.github.io.\n\nRegistration Category: Full Acc
 ess, Full Access Supporter\n\nLanguage Format: English Language\n\nSession
  Chair: Meng Zhang (Nanjing University of Science and Technology)
URL:https://asia.siggraph.org/2024/program/?id=papers_431&sess=sess118
END:VEVENT
END:VCALENDAR