The observation of deformation heterogeneities within polycrystalline materials has encouraged the development of mechanical models of crystal plasticity, taking into account the local behaviour of the microstructure [1]. Usually, the comparison between simulations with crystal plasticity models and experimental results are carried out by focusing on effective responses or microscopic surface data. The advent of 3D X-Ray Diffraction analysis, in synchrotron infrastructure, is a major opportunity for volume characterisation of polycrystalline materials [2]. These techniques allow the mapping of the microstructure, in terms of morphology, grain orientation but also elastic deformation, at resolutions (few hundreds of nanometres) and over thicknesses (few hundreds of micrometres) unachievable by more conventional techniques. Obtaining highly resolved experimental volume fields, witnessing the deformation of polycrystalline materials, opens the way to a full-field confrontation, with crystal plasticity simulations. The acceleration of crystal plasticity simulations by the Fast Fourier Transform (FFT) method offers the possibility of implementing a multiscale calibration procedure for crystal plasticity laws. The work presented here introduces an approach combining FFT simulations with machine learning, using Computer Vision, aimed at accelerating the transition between the scales of the simulation by a micromechanical super-resolution operation [3]. References: [1] Meric, L., Poubanne, P., and Cailletaud, G.. Single Crystal Modeling for Structural Calculations : Part 1—Model Presentation, Journal of Engineering Materials and Technology, 113(1): 162–170. 1991. [2] Reischig, P. and Ludwig, W. Three-dimensional reconstruction of intragranular strain and orientation in polycrystals by near-field X-ray diffraction, Current Opinion in Solid State and Materials Science, 24(5): 100851. 2020. [3] Dong, C., Loy, C. C., He, K., and Tang, X. Image Super-Resolution Using Deep Convolutional Networks. arXiv:1501.00092 [cs]. 2015.