Abstract: In this paper, we introduce MaeFuse, a novel autoencoder model designed for Infrared and Visible Image Fusion (IVIF). The existing approaches for image fusion often rely on training combined ...