A Review of Disentangled Representation Learning for Remote Sensing Data

Abstract

Representation learning is one of the core problems in machine learning research. The transition of input representations for machine learning algorithms from handcraft features, which dominated in the past, to the potential representations learned through deep neural networks nowadays has led to tremendous improvements in algorithm performance. However, the current representations are usually highly entangled, i.e., all information components of the input data are encoded into the same feature space, thus affecting each other and making it difficult to distinguish. Disentangled representation learning aims to learn a low-dimensional interpretable abstract representation that can identify and isolate different potential variables hidden in the high-dimensional observations. Disentangled representation learning can capture information about a single change factor and control it by the corresponding potential subspace, providing a robust representation for complex changes in the data. In this paper, we first introduce and analyze the current status of research on disentangled representation and its causal mechanisms and summarize three crucial properties of disentangled representation. Then, disentangled representation learning algorithms are classified into four categories and outlined in terms of both mathematical description and applicability. Subsequently, the loss functions and objective evaluation metrics commonly used in existing work on disentangled representation are classified. Finally, the paper summarizes representative applications of disentangled representation learning in the field of remote sensing and discusses its future development.