Abstract: Image fusion technology refers to the extraction and fusion of complementary information from different source images to generate an image with more information and sufficient support for subsequent high-level visual tasks. Infrared and Visible Image Fusion (IVIF) is an important branch of image fusion. In recent years, deep learning technology has shown good performance in various fields of visual computing. In particular, several deep learning-based IVIF technologies, such as autoencoders, convolutional neural networks, and generative adversarial networks, have been developed vigorously. Therefore, this paper summarizes and expounds on the methods, data sets, and evaluation indicators of the IVIF algorithm based on deep learning. Through a large number of experiments, qualitative and quantitative results are analyzed, and the performance of various IVIF algorithms based on deep learning is compared. Finally, some prospects for the future development of this field are discussed.