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:20251218T030655Z
LOCATION:Meeting Room S423+S424\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251216T113400
DTEND;TZID=Asia/Hong_Kong:20251216T114500
UID:siggraphasia_SIGGRAPH Asia 2025_sess117_papers_1761@linklings.com
SUMMARY:UltraZoom: Generating Gigapixel Images from Regular Photos
DESCRIPTION:Jingwei Ma, Vivek Jayaram, Brian Curless, Ira Kemelmacher-Shli
 zerman, and Steven Seitz (University of Washington)\n\nWe present UltraZoo
 m, a system for generating gigapixel-resolution images of objects from cas
 ually captured inputs, such as handheld phone photos. Given a full-shot im
 age (global, low-detail) and one or more close-ups (local, high-detail), U
 ltraZoom upscales the full image to match the fine detail and scale of the
  close-up examples. To achieve this, we construct a per-instance paired da
 taset from the close-ups and adapt a pretrained generative model to learn 
 object-specific low-to-high resolution mappings. At inference, we apply th
 e model in a sliding window fashion over the full image. Constructing thes
 e pairs is non-trivial: it requires registering the close-ups within the\n
 full image for scale estimation and degradation alignment. We introduce a 
 simple, robust method for achieving registration on arbitrary materials in
  casual, in-the-wild captures. Together, these components form a system th
 at enables seamless pan and zoom across the entire object, producing\ncons
 istent, photorealistic gigapixel imagery from minimal input. For full-reso
 lution results and code, visit our project page at ultra-zoom.github.io.\n
 \nRegistration Category: Full Access, Full Access Supporter\n\nSession Cha
 ir: Qiang Fu (King Abdullah University of Science and Technology (KAUST))\
 n\n
END:VEVENT
END:VCALENDAR