This thesis introduces novel methodologies for enhancing defect classification and characterization in advanced composite materials by leveraging state-of-the-art machine learning (ML), deep learning (DL), and federated ...