Capsule Network based Contrastive Learning of Unsupervised Visual Representations

TL;DR

This paper introduces Contrastive Capsule (CoCa), a novel Siamese Capsule Network for unsupervised visual representation learning, achieving high accuracy with significantly fewer parameters and FLOPs on CIFAR-10.

Contribution

The paper presents a new contrastive learning framework using Capsule Networks with a novel architecture, training, and testing algorithm for unsupervised image classification.

Findings

Achieved 70.50% top-1 accuracy on CIFAR-10

Model has 31 times fewer parameters than SOTA

Model has 71 times fewer FLOPs than SOTA

Abstract

Capsule Networks have shown tremendous advancement in the past decade, outperforming the traditional CNNs in various task due to it's equivariant properties. With the use of vector I/O which provides information of both magnitude and direction of an object or it's part, there lies an enormous possibility of using Capsule Networks in unsupervised learning environment for visual representation tasks such as multi class image classification. In this paper, we propose Contrastive Capsule (CoCa) Model which is a Siamese style Capsule Network using Contrastive loss with our novel architecture, training and testing algorithm. We evaluate the model on unsupervised image classification CIFAR-10 dataset and achieve a top-1 test accuracy of 70.50% and top-5 test accuracy of 98.10%. Due to our efficient architecture our model has 31 times less parameters and 71 times less FLOPs than the current SOTA in both supervised and unsupervised learning.