We demonstrate a deep-learning neural network (DNN) method for the measurement of molecular alignment by using the molecular-alignment-based cross-correlation polarization-gating frequency resolved optical gating (M-XFROG) technique. Our network has the capacity for direct measurement of molecular alignment from the FROG traces. In a proof-of-principle experiment, we have demonstrated our method in ${\mathrm{O}}_2$ molecules. With our method, the molecular alignment factor $⟨{\cos }^2\theta ⟩$(t) of ${\mathrm{O}}_2$, impulsively excited by a pump pulse, was directly reconstructed. The accuracy and validity of the reconstruction have been verified by comparison with the simulations based on experimental parameters.