Polymeric porous media have been investigated intensively in a wide range of fields due to their special characteristics, such as tunable pore sizes, shapes, densities, and large surface-to-volume ratio. A rigorous characterization of the microstructure of polymeric porous media is critical for understanding their properties and improving the performance for target applications. One promising method for the quantitative analysis of porous structures leverages the physics-based simulation of porous structures at the pore scale, which can be validated against real experimental microstructures, followed by extracting the structure-property relationships with data-driven algorithms, such as artificial neural networks. In this study, a Variational AutoEncoder (VAE) neural network is used to characterize the 3D structural information of porous materials with a low-dimensional latent space, which models the structure-property relationships. With the help of research data management platform Kadi4Mat (Karlsruhe Data Infrastructure for Materials Science, https://kadi.iam-cms.kit.edu), an efficient workflow is implemented, which can efficiently manage experimental and simulation data, to accelerate and streamline the research into microstructure materials.