iCaps: An Interpretable Classifier via Disentangled Capsule Networks

by   Dahuin Jung, et al.

We propose an interpretable Capsule Network, iCaps, for image classification. A capsule is a group of neurons nested inside each layer, and the one in the last layer is called a class capsule, which is a vector whose norm indicates a predicted probability for the class. Using the class capsule, existing Capsule Networks already provide some level of interpretability. However, there are two limitations which degrade its interpretability: 1) the class capsule also includes classification-irrelevant information, and 2) entities represented by the class capsule overlap. In this work, we address these two limitations using a novel class-supervised disentanglement algorithm and an additional regularizer, respectively. Through quantitative and qualitative evaluations on three datasets, we demonstrate that the resulting classifier, iCaps, provides a prediction along with clear rationales behind it with no performance degradation.



There are no comments yet.


page 11

page 12

page 13


Examining the Benefits of Capsule Neural Networks

Capsule networks are a recently developed class of neural networks that ...

Sparse Unsupervised Capsules Generalize Better

We show that unsupervised training of latent capsule layers using only t...

Capsule Networks for Protein Structure Classification and Prediction

Capsule Networks have great potential to tackle problems in structural b...

CapProNet: Deep Feature Learning via Orthogonal Projections onto Capsule Subspaces

In this paper, we formalize the idea behind capsule nets of using a caps...

Galaxy morphology prediction using capsule networks

Understanding morphological types of galaxies is a key parameter for stu...

Quantum Capsule Networks

Capsule networks, which incorporate the paradigms of connectionism and s...

DARCCC: Detecting Adversaries by Reconstruction from Class Conditional Capsules

We present a simple technique that allows capsule models to detect adver...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.