From Images to 3D Shape Attributes

David F. Fouhey, Abhinav Gupta, Andrew Zisserman

Research output: Contribution to journalArticlepeer-review


Our goal in this paper is to investigate properties of 3D shape that can be determined from a single image. We define 3D shape attributes - generic properties of the shape that capture curvature, contact and occupied space. Our first objective is to infer these 3D shape attributes from a single image. A second objective is to infer a 3D shape embedding - a low dimensional vector representing the 3D shape. We study how the 3D shape attributes and embedding can be obtained from a single image by training a Convolutional Neural Network (CNN) for this task. We start with synthetic images so that the contribution of various cues and nuisance parameters can be controlled. We then turn to real images and introduce a large scale image dataset of sculptures containing 143K images covering 2197 works from 242 artists. For the CNN trained on the sculpture dataset we show the following: (i) which regions of the imaged sculpture are used by the CNN to infer the 3D shape attributes; (ii) that the shape embedding can be used to match previously unseen sculptures largely independent of viewpoint; and (iii) that the 3D attributes generalize to images of other (non-sculpture) object classes.

Original languageEnglish (US)
Article number8194914
Pages (from-to)93-106
Number of pages14
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Issue number1
StatePublished - Jan 1 2019


  • 3D understanding
  • attributes
  • convolutional neural networks
  • shape perception

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Computational Theory and Mathematics
  • Artificial Intelligence
  • Applied Mathematics


Dive into the research topics of 'From Images to 3D Shape Attributes'. Together they form a unique fingerprint.

Cite this