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:20260114T163653Z
LOCATION:Meeting Room C4.9+C4.10\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T172000
DTEND;TZID=Australia/Melbourne:20231212T173000
UID:siggraphasia_SIGGRAPH Asia 2023_sess162_papers_204@linklings.com
SUMMARY:High-Fidelity and Real-Time Novel View Synthesis for Dynamic Scene
 s
DESCRIPTION:Haotong Lin (State Key Laboratory of CAD & CG, Zhejiang Univer
 sity); Sida Peng (Zhejiang University); and Zhen Xu, Tao Xie, Xingyi He, H
 ujun Bao, and Xiaowei Zhou (State Key Laboratory of CAD & CG, Zhejiang Uni
 versity)\n\nThis paper aims to tackle the challenge of dynamic view synthe
 sis from multi-view videos. The key observation is that while previous gri
 d-based methods offer consistent rendering, they fall short in capturing a
 ppearance details on a complex dynamic scene, a domain where multi-view im
 age-based methods demonstrate the opposite properties. To combine the best
  of two worlds, we introduce a hybrid scene representation that consists o
 f a grid-based geometry representation and a multi-view image-based appear
 ance representation. Specifically, the dynamic geometry is encoded as a 4D
  density function composed of spatiotemporal feature planes and a small ML
 P network, which globally models the scene structure and facilitates the r
 endering consistency. We represent the scene appearance by the original mu
 lti-view videos and a network that learns to predict the color of a 3D poi
 nt from image features, instead of totally memorizing the appearance with 
 networks, thereby naturally making the learning of networks easier. Our me
 thod is evaluated on five dynamic view synthesis datasets including DyNeRF
 , ZJU-MoCap, NHR, DNA-Rendering and ENeRF-Outdoor datasets. The results sh
 ow that the proposed representation exhibits state-of-the-art performance 
 in rendering quality and can be trained quickly, while being efficient for
  real-time rendering with a speed of 79.8 FPS for 512x512 images (ZJU-MoCa
 p dataset), on a single RTX 3090 GPU. The code is available at https://zju
 3dv.github.io/im4d.\n\nRegistration Category: Full Access\n\nSession Chair
 : Binh-Son Hua (Trinity College Dublin)\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_204&sess=sess162
END:VEVENT
END:VCALENDAR