DL-CapsNet: A Deep and Light Capsule Network
This work addresses efficiency issues in Capsule Networks for image classification, offering a more practical solution for handling datasets with many categories, though it appears incremental in nature.
The authors tackled the problem of high parameter complexity in Capsule Networks by proposing DL-CapsNet, a deep variant with a Capsule Summarization layer, which reduces parameters while maintaining high accuracy and enabling faster training and inference on complex datasets.
Capsule Network (CapsNet) is among the promising classifiers and a possible successor of the classifiers built based on Convolutional Neural Network (CNN). CapsNet is more accurate than CNNs in detecting images with overlapping categories and those with applied affine transformations. In this work, we propose a deep variant of CapsNet consisting of several capsule layers. In addition, we design the Capsule Summarization layer to reduce the complexity by reducing the number of parameters. DL-CapsNet, while being highly accurate, employs a small number of parameters and delivers faster training and inference. DL-CapsNet can process complex datasets with a high number of categories.