“…This class of methods mainly includes iterative-transform methods (developed from the Gerchberg-Saxton algorithm) [ 3 , 4 , 5 , 6 , 7 ], parametric methods (also known as model-based optimization algorithms or directly called phase diversity algorithms) [ 8 , 9 , 10 , 11 , 12 ], and deep learning methods [ 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 ]. Compared to other WFS methods, such as the Shack–Hartmann sensor [ 21 ], pyramid sensor [ 22 ], or curvature sensing [ 23 ], this class of WFS methods is particularly suitable for space applications [ 24 , 25 ].…”