Subspace Capsule Network

  • 2020-02-07 17:51:56
  • Marzieh Edraki, Nazanin Rahnavard, Mubarak Shah
  • 35


Convolutional neural networks (CNNs) have become a key asset to most offields in AI. Despite their successful performance, CNNs suffer from a majordrawback. They fail to capture the hierarchy of spatial relation amongdifferent parts of an entity. As a remedy to this problem, the idea of capsuleswas proposed by Hinton. In this paper, we propose the SubSpace Capsule Network(SCN) that exploits the idea of capsule networks to model possible variationsin the appearance or implicitly defined properties of an entity through a groupof capsule subspaces instead of simply grouping neurons to create capsules. Acapsule is created by projecting an input feature vector from a lower layeronto the capsule subspace using a learnable transformation. This transformationfinds the degree of alignment of the input with the properties modeled by thecapsule subspace. We show that SCN is a general capsule network that cansuccessfully be applied to both discriminative and generative models withoutincurring computational overhead compared to CNN during test time.Effectiveness of SCN is evaluated through a comprehensive set of experiments onsupervised image classification, semi-supervised image classification andhigh-resolution image generation tasks using the generative adversarial network(GAN) framework. SCN significantly improves the performance of the baselinemodels in all 3 tasks.


Quick Read (beta)

loading the full paper ...