Super-Resolution via Conditional Implicit Maximum Likelihood Estimation

Single-image super-resolution (SISR) is a canonical problem with diverse applications. Leading methods like SRGAN (Ledig et al., 2017) produce images that contain various artifacts, such as high-frequency noise, hallucinated colours and shape distortions, which adversely affect the realism of the result. In this paper, we propose an alternative approach based on an extension of the method of Implicit Maximum Likelihood Estimation (IMLE) (Li & Malik, 2018). We demonstrate greater effectiveness at noise reduction and preservation of the original colours and shapes, yielding more realistic super-resolved images. The problem of single-image super-resolution (SISR) aims to output a plausible high-resolution image that is consistent with a given low-resolution image. The key challenge arises from the fact that the problem is ill-posed - given the same low-resolution image, there are many different highresolution images that would be the same as the low-resolution image upon downsampling.