An Optimization view on Squash Function of CapsNet
DOI:
https://doi.org/10.54097/hset.v62i.10414Keywords:
Squash function, CapsNet, Capsule.Abstract
In CapsNet, a bounded measure of the modulus length of the feature is needed, so Squash function is used to compress the feature vector. This paper discusses the definition of Squash function, redefines Squash function based on the idea of information gain rate of decision tree, and constructs CapsNet model on this function. By testing on MNIST, Fashion-MNIST and Cifar-10 datasets, the experimental results show that the Squash function defined in this paper has better classification performance than the traditional Squash function in CapsNet model.
Downloads
References
Sabour S, Frosst N, Hinton G E. Dynamic routing between capsules. [C]//Advances in Neural Information Processing Systems. 2017: 3856-3866.
Wang, Dilin, Qiang Liu. An optimization view on dynamic routing between capsules. [C]//ICLR 2018.
Sabour, Sara, Nicholas Frosst, and G. Hinton. Matrix capsules with EM routing. [C]// 6th International Conference on Learning Representations, ICLR. 2018.
Rajasegaran, Jathushan, et al. DeepCaps: Going Deeper with Capsule Networks. [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019.
Siwei Chang, Jin Liu. Multi-Lane Capsule Network for Classifying Images With Complex Background[J]. IEEE ACCESS. 2020.8.
Ruiyang Shi, Lingfeng Niu , Ruizhi Zhou. Sparse CapsNet with explicit regularizer [J]. 2022.124,108486.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
