Search

Convolutional Neural Networks (CNNs) have been widely used in video super-resolution (VSR). Most existing VSR methods focus on how to utilize the information of multiple frames, while neglecting the feature correlations of the intermediate features, thus limiting the feature expression of the models. To address this problem, we propose a novel SAA network, that is, Scale-and-Attention-Aware Networks, to apply different attention to different temporal-length streams, while further exploring both spatial and channel attention on separate streams with a newly proposed Criss-Cross Channel Attention Module (C3AM). Experiments on public VSR datasets demonstrate the superiority of our method over other state-of-the-art methods in terms of both quantitative and qualitative metrics.

Video super-resolution (VSR) aims to utilize multiple low-resolution frames to generate a high-resolution prediction for each frame. In this process, inter- and intra-frames are the key sources for exploiting temporal and spatial information. However, there are a couple of limitations for existing VSR methods. First, optical flow is often used to establish one-on-one temporal correspondences. But flow estimation itself is error-prone and hence largely affects the ultimate recovery result. Second, similar patterns existing in natural images are rarely exploited for the VSR task. Motivated by these findings, we propose a temporal multi-correspondence aggregation strategy to leverage most similar patches across frames, and also a cross-scale nonlocal-correspondence aggregation scheme to explore self-similarity of images across scales. Based on these two novel modules, we build an effective multi-correspondence aggregation network (MuCAN) for VSR. Our method achieves state-of-the-art results on multiple benchmark datasets. Extensive experiments justify the effectiveness of our method.

This paper reviews the AIM 2019 challenge on extreme image super-resolution, the problem of restoring of rich details in a low resolution image. Compared to previous, this challenge focuses on an extreme upscaling factor, ×16, and employs the novel DIVerse 8K resolution (DIV8K) dataset. This report focuses on the proposed solutions and final results. The challenge had 2 tracks. The goal in Track 1 was to generate a super-resolution result with high fidelity, using the conventional PSNR as the primary metric to evaluate different methods. Track 2 instead focused on generating visually more pleasant super-resolution results, evaluated using subjective opinions. The two tracks had 71 and 52 registered participants, respectively, and 9 teams competed in the final testing phase. This report gauges the experimental protocol and baselines for the extreme image super-resolution task.