Springback is a crucial factor in cold stamping that causes geometric inaccuracy of the stamped component after removal of tools. This study, for the first time, presents a novel application of a Convolutional Neural Network (CNN) based surrogate model to predict the thinning and springback behaviours for cold stamping. Datasets were created based on two cold stamping case studies, i.e., a U-bending case and an outer car door panel stamping case. The datasets were then applied to train the CNN-based surrogate models. The results show that the surrogate models can achieve near indistinguishable full-field predictions in real-time when compared with the FE simulation results. The application of CNN in efficient springback prediction can be adopted in industrial settings to aid both conceptual and final component designs for designers without having manufacturing knowledge.