In capsule networks (CapsNets), the capsule is made up of collections of neurons. Their adjacent capsule layers are connected using routing-by-agreement mechanisms in an unsupervised way. The routing-by-agreement mechanisms have two main drawbacks: a) too many parameters and high computation complexity; b) the cluster distribution assumptions of these routing mechanisms may not hold in some complex real-world data. In this paper, we propose a novel Group Feedback Capsule Network (GF-CapsNet) which adopts a supervised routing strategy called group-routing. Compared with the previous routing strategies which globally transform each capsule, Group-routing equally splits capsules into groups where capsules locally share the same transformation weights, reducing routing parameters. To address the second drawback, we devise a distance network to directly predict capsules in a supervised way without making distribution assumptions. Our proposed group-routing captures local information of low-level capsules by group-wise transformation and supervisedly predicts high-level ones in a feedback way to address two drawbacks respectively. We conduct experiments on CIFAR-10/100 and SVHN datasets and the results show that our method can perform better against state-of-the-arts.

中文翻译：

---

组反馈胶囊网络

在胶囊网络（CapsNets）中，胶囊由神经元集合组成。使用协议路由机制，以无人监管的方式连接它们相邻的胶囊层。按协议路由机制有两个主要缺点：a）参数过多和计算复杂度高；b）这些路由机制的集群分布假设可能不适用于某些复杂的实际数据。在本文中，我们提出了一种新颖的组反馈胶囊网络（GF-CapsNet），该网络采用了一种称为组路由的监督路由策略。与以前的全局路由每个胶囊的路由策略相比，分组路由将胶囊平均分为几组，其中胶囊在本地共享相同的转换权重，从而减少了路由参数。为了解决第二个缺点，我们设计了一个距离网络，以一种有监督的方式直接预测胶囊，而无需做出分布假设。我们提出的分组路由通过分组变换捕获低级胶囊的本地信息，并以反馈方式监督预测高级胶囊，以分别解决两个缺点。我们在CIFAR-10 / 100和SVHN数据集上进行了实验，结果表明我们的方法可以针对最新技术进行更好的处理。