Autoencoders (AEs) are a class of artificial neural networks designed to learn efficient representations of input data. They consist of an encoder that maps input data to a lower-dimensional latent space, and a decoder that reconstructs the input data from this latent representation. In the context of image augmentation, autoencoders can be used to generate new, yet similar, images (Shorten and Khoshgoftaar in Journal of big data 6:1–48, 2019). By training an autoencoder on a dataset of images, the network learns to capture the underlying structure and variations within the data. Once trained, the encoder can be used to generate new latent codes, which can then be fed into the decoder to produce augmented images.