论文标题

生成胶囊模型的推理和学习

Inference and Learning for Generative Capsule Models

论文作者

Nazabal, Alfredo, Tsagkas, Nikolaos, Williams, Christopher K. I.

论文摘要

胶囊网络(参见例如Hinton等,2018)旨在编码有关对象及其部分之间关系的知识和理由。在本文中,我们为此类数据指定了一个生成模型,并得出了一种用于在场景中推断每个模型对象的转换以及观察到的部分对对象的分配的变异算法。我们基于各种期望最大化得出了对象模型的学习算法(Jordan等,1999)。我们还基于Fischler and Bolles(1981)的RANSAC方法研究了一种替代推理算法。我们将这些推理方法应用于(i)从正方形和三角形(“星座”)等多个几何对象生成的数据,以及(ii)来自面部零件模型的数据。 Kosiorek等人的最新工作。 (2019年)通过堆叠的胶囊自动编码器(SCAE)使用了摊销推理来解决此问题 - 我们的结果表明,我们在可以进行比较的地方(在星座数据上)大大优于它们。

Capsule networks (see e.g. Hinton et al., 2018) aim to encode knowledge of and reason about the relationship between an object and its parts. In this paper we specify a generative model for such data, and derive a variational algorithm for inferring the transformation of each model object in a scene, and the assignments of observed parts to the objects. We derive a learning algorithm for the object models, based on variational expectation maximization (Jordan et al., 1999). We also study an alternative inference algorithm based on the RANSAC method of Fischler and Bolles (1981). We apply these inference methods to (i) data generated from multiple geometric objects like squares and triangles ("constellations"), and (ii) data from a parts-based model of faces. Recent work by Kosiorek et al. (2019) has used amortized inference via stacked capsule autoencoders (SCAEs) to tackle this problem -- our results show that we significantly outperform them where we can make comparisons (on the constellations data).

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源