PSU_Yamamoto

A machine learning-based 3D reconstruction of highly porous cement samples cured in a microgravity environment is being pursued. The approach follows a deep learning-based framework based on the solid texture synthesis method, widely adopted in the computer graphics community. Preliminary results https://doi.org/10.2514/6.2023-2025 showed promising results, and with increased computational resources, larger exemplars higher resolution will be used for 3D reconstruction.