Latent process models have been widely applied to time series and spatial data which involve complex correlation structures. However, the existing approaches assume a known distributional property of the observations given the latent process. Furthermore, there seems to be no literature treating...