What Do Single-view 3D Reconstruction Networks Learn?

被引:273
|
作者
Tatarchenko, Maxim [1 ]
Richter, Stephan R. [2 ]
Ranftl, Rene [2 ]
Li, Zhuwen [2 ]
Koltun, Vladlen [2 ]
Brox, Thomas [1 ]
机构
[1] Univ Freiburg, Freiburg, Germany
[2] Intel Labs, Santa Clara, CA USA
关键词
SHAPE;
D O I
10.1109/CVPR.2019.00352
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Convolutional networks for single-view object reconstruction have shown impressive performance and have become a popular subject of research. All existing techniques are united by the idea of having an encoder-decoder network that performs non-trivial reasoning about the 3D structure of the output space. In this work, we set up two alternative approaches that perform image classification and retrieval respectively. These simple baselines yield better results than state-of-the-art methods, both qualitatively and quantitatively. We show that encoder-decoder methods are statistically indistinguishable from these baselines, thus indicating that the current state of the art in single-view object reconstruction does not actually perform reconstruction but image classification. We identify aspects of popular experimental procedures that elicit this behavior and discuss ways to improve the current state of research.
引用
收藏
页码:3400 / 3409
页数:10
相关论文
共 50 条
  • [41] PAniC-3D: Stylized Single-view 3D Reconstruction from Portraits of Anime Characters
    Chen, Shuhong
    Zhang, Kevin
    Shi, Yichun
    Wang, Heng
    Zhu, Yiheng
    Song, Guoxian
    An, Sizhe
    Kristjansson, Janus
    Yang, Xiao
    Zwicker, Matthias
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 21068 - 21077
  • [42] Deep Learning-Based Single-View Fluorescence Dose Reconstruction for 3D Dosimetry
    Nomura, Y.
    Ashraf, M. R.
    Xing, L.
    INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS, 2023, 117 (02): : S49 - S50
  • [43] 3D reconstruction of human bodies from single-view and multi-view images: A systematic review
    Correia, Helena A.
    Brito, Jose Henrique
    COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE, 2023, 239
  • [44] Single-view 3D reconstruction method based on neural radiation field occlusion optimization
    Chen, Zhijie
    Deng, Huiping
    Xiang, Sen
    Wu, Jin
    CHINESE JOURNAL OF LIQUID CRYSTALS AND DISPLAYS, 2024, 39 (10) : 1402 - 1410
  • [45] 3DCaricShop: A Dataset and A Baseline Method for Single-view 3D Caricature Face Reconstruction
    Qiu, Yuda
    Xu, Xiaojie
    Qiu, Lingteng
    Pan, Yan
    Wu, Yushuang
    Chen, Weikai
    Han, Xiaoguang
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 10231 - 10240
  • [46] Viewpoint-independent Single-view 3D Object Reconstruction using Reinforcement Learning
    Ito, Seiya
    Ju, Byeongjun
    Kaneko, Naoshi
    Sumi, Kazuhiko
    PROCEEDINGS OF THE 17TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS (VISAPP), VOL 5, 2022, : 811 - 819
  • [47] Few-Shot Single-View 3D Reconstruction with Memory Prior Contrastive Network
    Xing, Zhen
    Chen, Yijiang
    Ling, Zhixin
    Zhou, Xiangdong
    Xiang, Yu
    COMPUTER VISION - ECCV 2022, PT I, 2022, 13661 : 55 - 70
  • [48] Semi-supervised Single-View 3D Reconstruction via Prototype Shape Priors
    Xing, Zhen
    Li, Hengduo
    Wu, Zuxuan
    Jiang, Yu-Gang
    COMPUTER VISION - ECCV 2022, PT I, 2022, 13661 : 535 - 551
  • [49] Learning pose-invariant 3D object reconstruction from single-view images
    Peng, Bo
    Wang, Wei
    Dong, Jing
    Tan, Tieniu
    NEUROCOMPUTING, 2021, 423 : 407 - 418
  • [50] Pose Adaptive Dual Mixup for Few-Shot Single-View 3D Reconstruction
    Cheng, Ta-Ying
    Yang, Hsuan-Ru
    Trigoni, Niki
    Chen, Hwann-Tzong
    Liu, Tyng-Luh
    THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / THE TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2022, : 427 - 435