A LIGHTWEIGHT HYBRID REPRESENTATION FOR VIRTUAL COMPLEX SCENES

被引:0
|
作者
Shao, Shuai [1 ]
Jin, Xin [1 ]
机构
[1] Tsinghua Univ, Shenzhen Int Grad Sch, Shenzhen 518055, Peoples R China
关键词
Lightweight representation; shape modeling; hybrid rendering; real-time rendering; deep learning; FIELDS;
D O I
10.1109/ICIP49359.2023.10222844
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
While virtual scenes with complex geometries and high-resolution physically based rendering(PBR) material textures are capable of representing details, they often occupy a large amount of memory and require significant manual editing by artists. In this paper, we introduce a new framework for representing virtual scenes more efficiently. Our framework utilizes a lightweight hybrid representation that combines traditional polygon-based meshes and material textures with learned neural textures. By using simplified geometries and lower-resolution PBR material textures, our representation achieves a compact size while maintaining the freedom for user-friendly scene editing of materials and illuminations. Additionally, we have designed a hybrid rendering pipeline to render the lightweight representation with ray-traced visual quality at real-time frame rates. Our approach reduces memory requirements, eliminates the need for manual editing, and is easy to train. We demonstrate the effectiveness of our framework on a variety of virtual complex scenes.
引用
收藏
页码:2875 / 2879
页数:5
相关论文
共 50 条
  • [1] HCLT-YOLO: A Hybrid CNN and Lightweight Transformer Architecture for Object Detection in Complex Traffic Scenes
    Chen, Zhige
    Yang, Kai
    Wu, Yandong
    Yang, Hao
    Tang, Xiaolin
    IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, 2025, 74 (03) : 3681 - 3694
  • [2] Setting representation of natural panorama scenes and virtual view generation
    Yamada, Kunio
    Mochizuki, Kenji
    Naemura, Takeshi
    Aizawa, Kiyoharu
    Saito, Takahiro
    Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers, 2002, 56 (04): : 583 - 591
  • [3] Virtual view generation of natural panorama scenes by setting representation
    Yamada, K
    Mochizuki, K
    Naemura, T
    Aizawa, K
    Saito, T
    STEREOSCOPIC DISPLAYS AND VIRTUAL REALITY SYSTEMS IX, 2002, 4660 : 300 - 309
  • [4] Perceiving complex virtual scenes without visual guidance
    Jansson, G
    TOUCH IN VIRTUAL ENVIRONMENTS: HAPTICS AND THE DESIGN OF INTERACTIVE SYSTEMS, 2002, : 169 - 179
  • [5] Sensor Data Fusion for Textured Reconstruction and Virtual Representation of Alpine Scenes
    Haufel, Gisela
    Bulatov, Dimitri
    Solbrig, Peter
    EARTH RESOURCES AND ENVIRONMENTAL REMOTE SENSING/GIS APPLICATIONS VIII, 2017, 10428
  • [6] CNN-Based Lightweight Flame Detection Method in Complex Scenes
    Li X.
    Zhang D.
    Sun L.
    Xu Y.
    Moshi Shibie yu Rengong Zhineng/Pattern Recognition and Artificial Intelligence, 2021, 34 (05): : 415 - 422
  • [7] A Parametric Top-View Representation of Complex Road Scenes
    Wang, Ziyan
    Liu, Buyu
    Schulter, Samuel
    Chandraker, Manmohan
    2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 10317 - 10325
  • [8] Cortical representation of animate and inanimate objects in complex natural scenes
    Naselaris, Thomas
    Stansbury, Dustin E.
    Gallant, Jack L.
    JOURNAL OF PHYSIOLOGY-PARIS, 2012, 106 (5-6) : 239 - 249
  • [9] Rendering optical effects based on spectra representation in complex scenes
    Dong, Weiming
    ADVANCES IN COMPUTER GRAPHICS, PROCEEDINGS, 2006, 4035 : 719 - 726
  • [10] A hybrid license plate extraction method for complex scenes
    Le, Wangchao
    Li, Shaofa
    18TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, VOL 2, PROCEEDINGS, 2006, : 324 - +