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:20240214T070243Z
LOCATION:Meeting Room C4.9+C4.10\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T171000
DTEND;TZID=Australia/Melbourne:20231212T172000
UID:siggraphasia_SIGGRAPH Asia 2023_sess162_papers_672@linklings.com
SUMMARY:Inovis: Instant Novel-View Synthesis
DESCRIPTION:Technical Papers\n\nMathias Harrer and Linus Franke (Friedrich
 -Alexander-Universität Erlangen-Nürnberg); Laura Fink (Friedrich-Alexander
 -Universität Erlangen-Nürnberg, Fraunhofer IIS); and Marc Stamminger and T
 im Weyrich (Friedrich-Alexander-Universität Erlangen-Nürnberg)\n\nNovel-vi
 ew synthesis is an ill-posed problem in that it requires inference of prev
 iously unseen information. Recently, reviving the traditional field of ima
 ge-based rendering, neural methods proved particularly suitable for this i
 nterpolation/extrapolation task; however, they often require a-priori scen
 e-completeness or costly pre-processing steps and generally suffer from lo
 ng (scene-specific) training times. Our work draws from recent progress in
  neural spatio-temporal supersampling to enhance a state-of-the-art neural
  renderer’s ability to infer novel-view information at inference time. We 
 adapt a supersampling architecture [Xiao et al. 2020], which resamples pre
 viously rendered frames, to instead recombine nearby camera images in a mu
 lti-view dataset. These input frames are warped into a joint target frame,
  guided by the most recent (point-based) scene representation, followed by
  neural interpolation. The resulting architecture gains sufficient robustn
 ess to significantly improve transferability to previously unseen datasets
 . In particular, this enables novel applications for neural rendering wher
 e dynamically streamed content is directly incorporated in a (neural) imag
 e-based reconstruction of a scene. As we will show, our method reaches sta
 te-of-the-art performance when compared to previous works that rely on sta
 tic and sufficiently densely sampled scenes; in addition, we demonstrate o
 ur system's particular suitability for dynamically streamed content, where
  our approach is able to produce high-fidelity novel-view synthesis even w
 ith significantly fewer available frames than competing neural methods.\n\
 nRegistration Category: Full Access\n\nSession Chair: Binh-Son Hua (Trinit
 y College Dublin)
URL:https://asia.siggraph.org/2023/full-program?id=papers_672&sess=sess162
END:VEVENT
END:VCALENDAR