SimUSR: A Simple but Strong Baseline for Unsupervised Image Super-resolution
This provides a strong baseline for unsupervised super-resolution, particularly useful for researchers when multiple low-resolution images are available, though it is incremental as it simplifies existing approaches.
The paper tackles unsupervised image super-resolution without paired or ground truth high-resolution images by using multiple low-resolution images to create pseudo pairs, converting the problem into supervised learning. It shows that this simple method outperforms state-of-the-art unsupervised methods with faster runtime and reduces the gap to supervised models, achieving 1st in PSNR, 2nd in SSIM, and 13th in LPIPS in the NTIRE 2020 challenge.
In this paper, we tackle a fully unsupervised super-resolution problem, i.e., neither paired images nor ground truth HR images. We assume that low resolution (LR) images are relatively easy to collect compared to high resolution (HR) images. By allowing multiple LR images, we build a set of pseudo pairs by denoising and downsampling LR images and cast the original unsupervised problem into a supervised learning problem but in one level lower. Though this line of study is easy to think of and thus should have been investigated prior to any complicated unsupervised methods, surprisingly, there are currently none. Even more, we show that this simple method outperforms the state-of-the-art unsupervised method with a dramatically shorter latency at runtime, and significantly reduces the gap to the HR supervised models. We submitted our method in NTIRE 2020 super-resolution challenge and won 1st in PSNR, 2nd in SSIM, and 13th in LPIPS. This simple method should be used as the baseline to beat in the future, especially when multiple LR images are allowed during the training phase. However, even in the zero-shot condition, we argue that this method can serve as a useful baseline to see the gap between supervised and unsupervised frameworks.