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:20260817T171540Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241205T172800
DTEND;TZID=Asia/Tokyo:20241205T174000
UID:siggraphasia_SIGGRAPH Asia 2024_sess137_papers_920@linklings.com
SUMMARY:MV2MV: Multi-View Image Translation via View-Consistent Diffusion 
 Models
DESCRIPTION:Youcheng Cai, Runshi Li, and Ligang Liu (University of Science
  and Technology of China)\n\nImage translation has various applications in
  computer graphics and computer vision, aiming to transfer images from one
  domain to another. Thanks to the excellent generation capability of diffu
 sion models, recent single-view image translation methods achieve realisti
 c results. However, directly applying diffusion models for multi-view imag
 e translation remains challenging for two major obstacles: the need for pa
 ired training data and the limited view consistency. To overcome the obsta
 cles, we present a unified multi-view image to multi-view image translatio
 n framework based on diffusion models, called MV2MV. Firstly, we propose a
  novel self-supervised training strategy that exploits the success of off-
 the-shelf single-view image translators and the 3D Gaussian Splatting (3DG
 S) technique to generate pseudo ground truths as supervisory signals, lead
 ing to enhanced consistency and fine details. Additionally, we propose a l
 atent multi-view consistency block, which utilizes the latent-3DGS as the 
 underlying 3D representation to facilitate information exchange across mul
 ti-view images and inject 3D prior into the diffusion model to enforce con
 sistency. Finally, our approach simultaneously optimizes the diffusion mod
 el and 3DGS to achieve a better trade-off between consistency and realism.
  Extensive experiments across various translation tasks demonstrate that M
 V2MV outperforms task-specific specialists in both quantitative and qualit
 ative.\n\nRegistration Category: Full Access, Full Access Supporter\n\nLan
 guage Format: English Language\n\nSession Chair: Michael Rubinstein (Googl
 e)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_920&sess=sess137
END:VEVENT
END:VCALENDAR