Capsule Network is Not More Robust than Convolutional Network
This work addresses the robustness comparison problem for researchers in neural network design, showing incremental improvements by identifying and integrating beneficial CapsNet components into ConvNets.
The paper challenges the belief that Capsule Networks are more robust than Convolutional Networks by conducting comprehensive ablation studies on three robustness types, revealing that key CapsNet components like dynamic routing can harm robustness, and proposes enhanced ConvNets that achieve better robustness than CapsNet.
The Capsule Network is widely believed to be more robust than Convolutional Networks. However, there are no comprehensive comparisons between these two networks, and it is also unknown which components in the CapsNet affect its robustness. In this paper, we first carefully examine the special designs in CapsNet that differ from that of a ConvNet commonly used for image classification. The examination reveals five major new/different components in CapsNet: a transformation process, a dynamic routing layer, a squashing function, a marginal loss other than cross-entropy loss, and an additional class-conditional reconstruction loss for regularization. Along with these major differences, we conduct comprehensive ablation studies on three kinds of robustness, including affine transformation, overlapping digits, and semantic representation. The study reveals that some designs, which are thought critical to CapsNet, actually can harm its robustness, i.e., the dynamic routing layer and the transformation process, while others are beneficial for the robustness. Based on these findings, we propose enhanced ConvNets simply by introducing the essential components behind the CapsNet's success. The proposed simple ConvNets can achieve better robustness than the CapsNet.