BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240214T070241Z
LOCATION:Darling Harbour Theatre\, Level 2 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T093000
DTEND;TZID=Australia/Melbourne:20231212T124500
UID:siggraphasia_SIGGRAPH Asia 2023_sess209_papers_644@linklings.com
SUMMARY:FuseSR: Super Resolution for Real-time Rendering through Efficient
  Multi-resolution Fusion
DESCRIPTION:Technical Papers\n\nZhihua Zhong (State Key Lab of CAD&CG, Zhe
 jiang University; Zhejiang University City College); Jingsen Zhu (State Ke
 y Lab of CAD&CG, Zhejiang University); Yuxin Dai (Zhejiang A&F University)
 ; Chuankun Zheng (State Key Lab of CAD&CG, Zhejiang University); Guanlin C
 hen (Zhejiang University City College); Yuchi Huo (Zhejiang Lab; State Key
  Lab of CAD&CG, Zhejiang University); and Hujun Bao and Rui Wang (State Ke
 y Lab of CAD&CG, Zhejiang University)\n\nThe workload of real-time renderi
 ng is steeply increasing as the demand for high resolution, high refresh r
 ates, and high realism rises, overwhelming most graphics cards. To mitigat
 e this problem, one of the most popular solutions is to render images at a
  low resolution to reduce rendering overhead, and then manage to accuratel
 y upsample the low-resolution rendered image to the target resolution, a.k
 .a. super-resolution techniques. Most existing methods focus on exploiting
  information from low-resolution inputs, such as historical frames. The ab
 sence of high frequency details in those LR inputs makes them hard to reco
 ver fine details in their high-resolution predictions. In this paper, we p
 ropose an efficient and effective super-resolution method that predicts hi
 gh-quality upsampled reconstructions utilizing low-cost high-resolution au
 xiliary G-Buffers as additional input. With LR images and HR G-buffers as 
 input, the network requires to align and fuse features at multi resolution
  levels. We introduce an efficient and effective H-Net architecture to sol
 ve this problem and significantly reduce rendering overhead without notice
 able quality deterioration. Experiments show that our method is able to pr
 oduce temporally consistent reconstructions in $4 \times 4$ and even chall
 enging $8 \times 8$ upsampling cases at 4K resolution with real-time perfo
 rmance, with substantially improved quality and significant performance bo
 ost compared to existing works.\n\nRegistration Category: Full Access, Enh
 anced Access, Trade Exhibitor, Experience Hall Exhibitor
URL:https://asia.siggraph.org/2023/full-program?id=papers_644&sess=sess209
END:VEVENT
END:VCALENDAR