We present a novel approach to computer-generated holography (CGH) using the Generalized Van Cittert-Zernike Schell propagator (GVS), using incoherent light emitted from the screen of a mobile phone, and a spatial light modulator (SLM). In this approach, we solve an inverse problem by jointly optimizing an RGB image displayed on the mobile phone screen and a phase pattern displayed on the SLM to produce the desired 3D scene at a given propagation distance. For this purpose, we use the stochastic gradient descent algorithm with the Adam optimizer and the GVS propagator to model the diffraction of partially spatially coherent light passing through the optical setup, which requires only a single forward propagation. This propagator uses the Van Cittert-Zernike theorem to compute the complex coherence factor from a light source reaching the object plane, and then applies the generalized Schell{\textquoteright}s theorem to compute the diffraction pattern at a given propagation distance, not limited to the far field. We present the results of the optimization for two target images at two different propagation distances, as well as for a much larger focal stack corresponding to a 3D object. Through numerical simulations and optical experiments, we show that our method is orders of magnitude faster and highly accurate compared to the state of the art.