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:20241204T145900
DTEND;TZID=Asia/Tokyo:20241204T151300
UID:siggraphasia_SIGGRAPH Asia 2024_sess118_papers_671@linklings.com
SUMMARY:FabricDiffusion: High-Fidelity Texture Transfer for 3D Garments Ge
 neration from In-The-Wild Images
DESCRIPTION:Technical Papers\n\nCheng Zhang (Carnegie Mellon University, T
 exas A&M University); Yuanhao Wang and Francisco Vicente (Carnegie Mellon 
 University); Chenglei Wu, Jinlong Yang, and Thabo Beeler (Google Inc.); an
 d Fernando De la Torre (Carnegie Mellon University)\n\nWe introduce Fabric
 Diffusion, a method for transferring fabric textures from a single clothin
 g image to 3D garments of arbitrary shapes. Existing approaches typically 
 synthesize textures on the garment surface through 2D-to-3D texture mappin
 g or depth-aware inpainting via generative models. Unfortunately, these me
 thods often struggle to capture and preserve texture details, particularly
  due to challenging occlusions, distortions, or poses in the input image. 
 Inspired by the observation that in the fashion industry, most garments ar
 e constructed by stitching sewing patterns with flat, repeatable textures,
  we cast the task of clothing texture transfer as extracting distortion-fr
 ee, tileable texture materials that are subsequently mapped onto the UV sp
 ace of the garment. Building upon this insight, we train a denoising diffu
 sion model with a large-scale synthetic dataset to rectify distortions in 
 the input texture image. This process yields a flat texture map that enabl
 es a tight coupling with existing Physically-Based Rendering (PBR) materia
 l generation pipelines, allowing for realistic relighting of the garment u
 nder various lighting conditions. We show that FabricDiffusion can transfe
 r various features from a single clothing image including texture patterns
 , material properties, and detailed prints and logos. Extensive experiment
 s demonstrate that our model significantly outperforms state-to-the-art me
 thods on both synthetic data and real-world, in-the-wild clothing images w
 hile generalizing to unseen textures and garment shapes.\n\nRegistration C
 ategory: Full Access, Full Access Supporter\n\nLanguage Format: English La
 nguage\n\nSession Chair: Meng Zhang (Nanjing University of Science and Tec
 hnology)
URL:https://asia.siggraph.org/2024/program/?id=papers_671&sess=sess118
END:VEVENT
END:VCALENDAR