This paper reviews the AIM 2020 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor x4 based on a set of prior examples of low and corresponding high resolution images. The goal is to devise a network that reduces one or several aspects such as runtime, parameter count, FLOPs, activations, and memory consumption while at least maintaining PSNR of MSRResNet. The track had 150 registered participants, and 25 teams submitted the final results. They gauge the state-of-the-art in efficient single image super-resolution.
Based on the great success of deterministic learning, to interactively control the output effects has attracted increasingly attention in the image restoration field. The goal is to generate continuous restored images by adjusting a controlling coefficient. Existing methods are restricted in realizing smooth transition between two objectives, while the real input images may contain different kinds of degradations. To make a step forward, we present a new problem called multi-dimension (MD) modulation, which aims at modulating output effects across multiple degradation types and levels. Compared with the previous single-dimension (SD) modulation, the MD task has three distinct properties, namely joint modulation, zero starting point and unbalanced learning. These obstacles motivate us to propose the first MD modulation framework -- CResMD with newly introduced controllable residual connections. Specifically, we add a controlling variable on the conventional residual connection to allow a weighted summation of input and residual. The exact values of these weights are generated by a condition network. We further propose a new data sampling strategy based on beta distribution to balance different degradation types and levels. With the corrupted image and the degradation information as inputs, the network could output the corresponding restored image. By tweaking the condition vector, users are free to control the output effects in MD space at test time. Extensive experiments demonstrate that the proposed CResMD could achieve excellent performance on both SD and MD modulation tasks.
In image restoration tasks, like denoising and super resolution, continual modulation of restoration levels is of great importance for real-world applications, but has failed most of existing deep learning based image restoration methods. Learning from discrete and fixed restoration levels, deep models cannot be easily generalized to data of continuous and unseen levels. This topic is rarely touched in literature, due to the difficulty of modulating well-trained models with certain hyper-parameters. We make a step forward by proposing a unified CNN framework that consists of few additional parameters than a single-level model yet could handle arbitrary restoration levels between a start and an end level. The additional module, namely AdaFM layer, performs channel-wise feature modification, and can adapt a model to another restoration level with high accuracy. By simply tweaking an interpolation coefficient, the intermediate model - AdaFM-Net could generate smooth and continuous restoration effects without artifacts. Extensive experiments on three image restoration tasks demonstrate the effectiveness of both model training and modulation testing. Besides, we carefully investigate the properties of AdaFM layers, providing a detailed guidance on the usage of the proposed method.