Self-Supervised Learning (SSL) has emerged as a powerful technique for learning meaningful representations from data without requiring labeled samples. This project focuses on leveraging self-supervised learning for Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR) to address critical challenges such as data scarcity and domain adaptation. Traditional SAR ATR methods struggle due to variations in pose, occlusion, and clutter, making it difficult to achieve high accuracy in real-world scenarios. This research integrates machine learning, computer vision, and neuroscience, aligning with ARNI’s Theme 1 (Continual Learning), Theme 4 (Neural Mechanisms of Intelligence), and Theme 5 (Language and Vision). The integration of neuroscience principles, such as hierarchical learning and viewpoint invariance, strengthens SAR ATR models’ generalization abilities. The research team consists of Dr. Dewan Fahim Noor from Tuskegee University as the principal investigator and Dr. Kenneth D. Miller from Columbia University as the co-investigator. The project is being carried out by graduate student Md Al Siam, who has been leading the implementation and experimental validation efforts.