Learning Signed Distance Field for Multi-view Surface Reconstruction

被引:42
|
作者
Zhang, Jingyang [1 ]
Yao, Yao [1 ]
Quan, Long [1 ]
机构
[1] Hong Kong Univ Sci & Technol, Hong Kong, Peoples R China
关键词
STEREO;
D O I
10.1109/ICCV48922.2021.00646
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recent works on implicit neural representations have shown promising results for multi-view surface reconstruction. However, most approaches are limited to relatively simple geometries and usually require clean object masks for reconstructing complex and concave objects. In this work, we introduce a novel neural surface reconstruction framework that leverages the knowledge of stereo matching and feature consistency to optimize the implicit surface representation. More specifically, we apply a signed distance field (SDF) and a surface light field to represent the scene geometry and appearance respectively. The SDF is directly supervised by geometry from stereo matching, and is refined by optimizing the multi-view feature consistency and the fidelity of rendered images. Our method is able to improve the robustness of geometry estimation and support reconstruction of complex scene topologies. Extensive experiments have been conducted on DTU, EPFL and Tanks and Temples datasets. Compared to previous state-of-the-art methods, our method achieves better mesh reconstruction in wide open scenes without masks as input.
引用
收藏
页码:6505 / 6514
页数:10
相关论文
共 50 条
  • [31] Multi-view learning with Universum
    Wang, Zhe
    Zhu, Yujin
    Liu, Wenwen
    Chen, Zhihua
    Gao, Daqi
    KNOWLEDGE-BASED SYSTEMS, 2014, 70 : 376 - 391
  • [32] High Resolution Surface Reconstruction from Multi-view Aerial Imagery
    Calakli, Fatih
    Ulusoy, Ali O.
    Restrepo, Maria I.
    Taubin, Gabriel
    Mundy, Joseph L.
    SECOND JOINT 3DIM/3DPVT CONFERENCE: 3D IMAGING, MODELING, PROCESSING, VISUALIZATION & TRANSMISSION (3DIMPVT 2012), 2012, : 25 - 32
  • [33] Mahalanobis Distance-Based Multi-view Optimal Transport for Multi-view Crowd Localization
    Zhang, Qi
    Zhang, Kaiyi
    Chan, Antoni B.
    Huang, Hui
    COMPUTER VISION-ECCV 2024, PT IV, 2025, 15062 : 19 - 36
  • [34] AUDIO-VISUAL SPEAKER IDENTIFICATION WITH MULTI-VIEW DISTANCE METRIC LEARNING
    Zheng, Haomian
    Wang, Meng
    Li, Zhu
    2010 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, 2010, : 4561 - 4564
  • [35] RayDF: Neural Ray-surface Distance Fields with Multi-view Consistency
    Liu, Zhuoman
    Yang, Bo
    Luximon, Yan
    Kumar, Ajay
    Li, Jinxi
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 36 (NEURIPS 2023), 2023,
  • [36] Multi-View Multi-Instance Learning Based on Joint Sparse Representation and Multi-View Dictionary Learning
    Li, Bing
    Yuan, Chunfeng
    Xiong, Weihua
    Hu, Weiming
    Peng, Houwen
    Ding, Xinmiao
    Maybank, Steve
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2017, 39 (12) : 2554 - 2560
  • [37] A regularized point-to-manifold distance metric for multi-view multi-manifold learning
    Aeini, Faraein
    Moghadam, Amir Masoud Eftekhari
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2019, 82 : 85 - 95
  • [38] MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo
    Chen, Anpei
    Xu, Zexiang
    Zhao, Fuqiang
    Zhang, Xiaoshuai
    Xiang, Fanbo
    Yu, Jingyi
    Su, Hao
    2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, : 14104 - 14113
  • [39] Ref-NeuS: Ambiguity-Reduced Neural Implicit Surface Learning for Multi-View Reconstruction with Reflection
    Ge, Wenhang
    Hu, Tao
    Zhao, Haoyu
    Liu, Shu
    Chen, Ying-Cong
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION, ICCV, 2023, : 4228 - 4237
  • [40] NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction
    Wang, Peng
    Liu, Lingjie
    Liu, Yuan
    Theobalt, Christian
    Komura, Taku
    Wang, Wenping
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 34 (NEURIPS 2021), 2021, 34