Equivariant neural networks, whose hidden features transform according to representations of a group G acting on the data, exhibit training efficiency and an improved generalisation performance. In this work, we extend group invariant and equivariant representation learning to the field of unsupervised deep learning. We propose a general learning strategy based on an encoder-decoder framework in which the latent representation is disentangled in an invariant term and an equivariant group action component. The key idea is that the network learns the group action on the data space and thus is able to solve the reconstruction task from an invariant data representation, hence avoiding the necessity of ad-hoc group-specific implementations. We derive the necessary conditions on the equivariant encoder, and we present a construction valid for any G, both discrete and continuous. We describe explicitly our construction for rotations, translations and permutations. We test the validity and the robustness of our approach in a variety of experiments with diverse data types employing different network architectures.

本文针对深度学习的无监督学习，将群不变和群等变表示学习扩展到了该领域。我们提出了一种基于编码器-解码器框架的通用学习策略，其中潜在表示被分为不变项和等变群作用项。在利用预测适当的群作用来对齐输入和输出姿势以解决重建任务时，网络可以学习将数据编码和解码为群不变表示。我们导出依变编码器的必要条件，并针对旋转，平移和置换明确描述了我们的构造。我们在不同网络架构下使用不同数据类型进行各种实验，测试了我们方法的有效性和鲁棒性。

无监督学习群不变和等变表示