3D Morphable Models as Spatial Transformer Networks


Bas A. , Huber P., Smith W. A. P. , Awais M., Kittler J.

16th IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 22 - 29 October 2017, pp.895-903 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/iccvw.2017.110
  • City: Venice
  • Country: Italy
  • Page Numbers: pp.895-903

Abstract

In this paper, we show how a 3D Morphable Model (i.e. a statistical model of the 3D shape of a class of objects such as faces) can be used to spatially transform input data as a module (a 3DMM-STN) within a convolutional neural network. This is an extension of the original spatial transformer network in that we are able to interpret and normalise 3D pose changes and self-occlusions. The trained localisation part of the network is independently useful since it learns to fit a 3D morphable model to a single image. We show that the localiser can be trained using only simple geometric loss functions on a relatively small dataset yet is able to perform robust normalisation on highly uncontrolled images including occlusion, self-occlusion and large pose changes.