Amplifying the imaging power of ground-based telescopes with space-based telescopes data and generative AI

The Enhanced Galaxy Imaging Dataset provides a collection of 63,202 galaxy images from the DESI Legacy Survey that have been processed using a custom generative AI model to match the image quality of space-based telescopes. The dataset combines the wide-field imaging capabilities of the Dark Energy Camera (DECam) with AI-driven image enhancement trained on high-resolution space telescope data. The "catalog" folder includes two main components: the "enhanced_galaxies" directory containing 256×256 AI-enhanced galaxy images, and the "galaxy_comparisons" directory featuring side-by-side collages of the original ground-based, AI-enhanced, and corresponding space-based images for qualitative evaluation.The "train" folder includes the concatenated samples used in training the cGAN model. Additionally, the "galaxy_enhancer_cgan.h5" is the trained model with learned weights and biases, enabling users to reproduce the end-to-end pipeline.