论文标题
在变换胶囊网络中的模棱两可胶囊之间的迭代协作路由
Iterative collaborative routing among equivariant capsules for transformation-robust capsule networks
论文作者
论文摘要
对于执行图像分类的机器学习模型来说，转换式运动是一个重要功能。许多方法旨在通过使用数据增强策略将该属性授予模型，而通过使用Equivariant模型获得更正式的保证。我们认识到，组成或部分整体结构也是图像的重要方面，必须考虑用于构建转换模型。因此，我们提出了一个胶囊网络模型，该模型同时是均等和组成性感知的。我们的胶囊网络模型的模棱两可来自在精心挑选的新型架构中使用模棱两可的卷积。组成性的意识来自我们提出的小说，基于图的路由算法的使用，称为迭代协作路由（ICR）。 ICR是我们贡献的核心，加权基于其最近邻居的程度中心的迭代平均分数对胶囊的预测。关于FashionMnist，CIFAR-10和CIFAR-100的转换图像分类的实验表明，我们使用ICR优于卷积和胶囊基准的模型以实现最新的性能。
Transformation-robustness is an important feature for machine learning models that perform image classification. Many methods aim to bestow this property to models by the use of data augmentation strategies, while more formal guarantees are obtained via the use of equivariant models. We recognise that compositional, or part-whole structure is also an important aspect of images that has to be considered for building transformation-robust models. Thus, we propose a capsule network model that is, at once, equivariant and compositionality-aware. Equivariance of our capsule network model comes from the use of equivariant convolutions in a carefully-chosen novel architecture. The awareness of compositionality comes from the use of our proposed novel, iterative, graph-based routing algorithm, termed Iterative collaborative routing (ICR). ICR, the core of our contribution, weights the predictions made for capsules based on an iteratively averaged score of the degree-centralities of its nearest neighbours. Experiments on transformed image classification on FashionMNIST, CIFAR-10, and CIFAR-100 show that our model that uses ICR outperforms convolutional and capsule baselines to achieve state-of-the-art performance.
