Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Iterative Collaborative Routing among Equivariant Capsules for Transformation-Robust Capsule Networks


Affiliations
1 Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, India
     

   Subscribe/Renew Journal


Transformation-robustness is an important feature for machine learning models that perform image classification. Many methods aim to bestow this property to models by the use of data augmentation strategies, while more formal guarantees are obtained via the use of equivariant models. We recognise that compositional, or part-whole structure is also an important aspect of images that has to be considered for building transformation-robust models. Thus, we propose a capsule network model that is, at once, equivariant and compositionality aware. Equivariance of our capsule network model comes from the use of equivariant convolutions in a carefully-chosen novel architecture. The awareness of compositionality comes from the use of our proposed novel, iterative, graph-based routing algorithm, termed Iterative collaborative routing (ICR). ICR, the core of our contribution, weights the predictions made for capsules based on an iteratively averaged score of the degree-centralities of its nearest neighbours. Experiments on transformed image classification on FashionMNIST, CIFAR-10, and CIFAR-100 show that our model that uses ICR outperforms convolutional and capsule baselines to achieve state-of-the-art performance.

Keywords

Equivariance, Transformation Robustness, Capsule Network, Image Classification, Deep Learning.
Subscription Login to verify subscription
User
Notifications
Font Size


  • Iterative Collaborative Routing among Equivariant Capsules for Transformation-Robust Capsule Networks

Abstract Views: 280  |  PDF Views: 1

Authors

Sai Raam Venkataraman
Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, India
S. Balasubramanian
Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, India
R. Raghunatha Sarma
Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, India

Abstract


Transformation-robustness is an important feature for machine learning models that perform image classification. Many methods aim to bestow this property to models by the use of data augmentation strategies, while more formal guarantees are obtained via the use of equivariant models. We recognise that compositional, or part-whole structure is also an important aspect of images that has to be considered for building transformation-robust models. Thus, we propose a capsule network model that is, at once, equivariant and compositionality aware. Equivariance of our capsule network model comes from the use of equivariant convolutions in a carefully-chosen novel architecture. The awareness of compositionality comes from the use of our proposed novel, iterative, graph-based routing algorithm, termed Iterative collaborative routing (ICR). ICR, the core of our contribution, weights the predictions made for capsules based on an iteratively averaged score of the degree-centralities of its nearest neighbours. Experiments on transformed image classification on FashionMNIST, CIFAR-10, and CIFAR-100 show that our model that uses ICR outperforms convolutional and capsule baselines to achieve state-of-the-art performance.

Keywords


Equivariance, Transformation Robustness, Capsule Network, Image Classification, Deep Learning.

References