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
  • [31] Sym3DNet: Symmetric 3D Prior Network for Single-View 3D Reconstruction
    Siddique, Ashraf
    Lee, Seungkyu
    SENSORS, 2022, 22 (02)
  • [32] Single-View 3D Face Reconstruction via Cross-View Consistency Constraints
    Zhong Y.
    Pei Y.
    Li P.
    Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics, 2024, 36 (04): : 543 - 551
  • [33] Perspective Transformer Nets: Learning Single-View 3D Object Reconstruction without 3D Supervision
    Yan, Xinchen
    Yang, Jimei
    Yumer, Ersin
    Guo, Yijie
    Lee, Honglak
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 29 (NIPS 2016), 2016, 29
  • [34] Enhancing single-view 3D mesh reconstruction with the aid of implicit surface learning
    Fahim, George
    Amin, Khalid
    Zarif, Sameh
    IMAGE AND VISION COMPUTING, 2022, 119
  • [35] Single-view 3D Scene Reconstruction with High-fidelity Shape and Texture
    Chen, Yixin
    Ni, Junfeng
    Jiang, Nan
    Zhang, Yaowei
    Zhu, Yixin
    Huang, Siyuan
    2024 INTERNATIONAL CONFERENCE IN 3D VISION, 3DV 2024, 2024, : 1456 - 1467
  • [36] LIST: Learning Implicitly from Spatial Transformers for Single-View 3D Reconstruction
    Arshad, Mohammad Samiul
    Beksi, William J.
    2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2023), 2023, : 9287 - 9296
  • [37] Single-View 3D Reconstruction via Differentiable Rendering and Inverse Procedural Modeling
    Garifullin, Albert
    Maiorov, Nikolay
    Frolov, Vladimir
    Voloboy, Alexey
    SYMMETRY-BASEL, 2024, 16 (02):
  • [38] A new single-view 3D pantograph reconstruction aided by prior CAD model
    Sun, Tiecheng
    Liu, Guanghui
    Peng, Jianping
    Meng, Fanman
    Liu, Shuaicheng
    Zhu, Shuyuan
    MEASUREMENT, 2021, 181
  • [39] Single-View 3D Reconstruction Based on Gradient-Applied Weighted Loss
    Kim, Taehyeon
    Lee, Jiho
    Lee, Kyung-Taek
    Choe, Yoonsik
    JOURNAL OF ELECTRICAL ENGINEERING & TECHNOLOGY, 2024, 19 (07) : 4523 - 4535
  • [40] Xtru3d: Single-view 3d object reconstruction from color and depth data
    Rodríguez-Jiménez, Silvia, 1600, Springer Verlag (458):