Most of the existing works in supervised spatio-temporal video super-resolution (STVSR) heavily rely on a large-scale external dataset consisting of paired low-resolution low-frame rate (LR-LFR)and high-resolution high-frame-rate (HR-HFR) videos. Despite their remarkable performance, these methods make a prior assumption that the low-resolution video is obtained by down-scaling the high-resolution video using a known degradation kernel, which does not hold in practical settings. Another problem with these methods is that they cannot exploit instance-specific internal information of video at testing time. Recently, deep internal learning approaches have gained attention due to their ability to utilize the instance-specific statistics of a video. However, these methods have a large inference time as they require thousands of gradient updates to learn the intrinsic structure of the data. In this work, we presentAdaptiveVideoSuper-Resolution (Ada-VSR) which leverages external, as well as internal, information through meta-transfer learning and internal learning, respectively. Specifically, meta-learning is employed to obtain adaptive parameters, using a large-scale external dataset, that can adapt quickly to the novel condition (degradation model) of the given test video during the internal learning task, thereby exploiting external and internal information of a video for super-resolution. The model trained using our approach can quickly adapt to a specific video condition with only a few gradient updates, which reduces the inference time significantly. Extensive experiments on standard datasets demonstrate that our method performs favorably against various state-of-the-art approaches.
Sliding Window Recurrent Network for Efficient Video Super-Resolution
Uppsala University（乌普萨拉大学）；Northeastern University（东北大学）
视频超分辨率（VSR）是从一系列低分辨率输入恢复高分辨率帧的任务。与单帧图像超分辨率不同，VSR可以利用帧间的时间信息重构出更多细节的图像。近年来，随着卷积神经网络（CNN）的迅速发展，VSR问题受到越来越多的关注，许多基于CNN的方法取得了显著的效果。然而，由于计算资源和运行时间的限制，只有少数VSR方法可以应用于现实世界的移动设备。在本文中，我们提出了一个\textit{Sliding Window based Recurrent Network}（SWRN），它可以是实时推理，同时仍然实现优越的性能。具体来说，我们注意到视频帧应该具有有助于恢复细节的空间和时间关系，而关键是如何提取和聚合信息。针对这个问题，我们输入三个相邻的帧，并利用一个隐藏状态来循环存储和更新重要的时间信息。我们在REDS数据集上的实验表明，该方法可以很好地适应移动设备，并产生视觉上令人愉快的结果。
FCA2: Frame Compression-Aware Autoencoder for Modular and Fast Compressed Video Super-Resolution
State Key Laboratory of Integrated Services Networks, School of Electronic Engineering, Xidian University（信息服务网络国家重点实验室，电子工程学院，西安电子科技大学）；Key Laboratory of Collaborative Intelligence Systems, Ministry of Education, Xidian University（协同智能系统重点实验室，教育部，西安电子科技大学）；School of Electronic Engineering, Xidian University（电子工程学院，西安电子科技大学）；Chongqing Key Laboratory of Image Cognition, Chongqing University of Posts and Telecommunications（重庆图像认知重点实验室，重庆邮电大学）
Compressed-Domain-Aware Online Video Super-Resolution
School of Information and Electronics, Beijing Institute of Technology, Beijing, China（信息与电子学院，北京理工大学，北京，中国）；Terahertz Science and Application Center, Beijing Institute of Technology, Zhuhai, Guangdong, China（太赫兹科学与应用中心，北京理工大学，珠海，广东，中国）
MambaVSR: Content-Aware Scanning State Space Model for Video Super-Resolution
Institute of Information Science, Beijing Jiaotong University（北京交通大学信息科学学院）；Visual Intelligence + X International Cooperation Joint Laboratory of MOE, Beijing Jiaotong University（教育部视觉智能+X国际合作联合实验室）；School of Computer and Communication Engineering, University of Science and Technology Beijing（北京科技大学计算机与通信工程学院）
FDAN: Flow-guided Deformable Alignment Network for Video Super-Resolution
Center for Research on Intelligent Perception and Computing (CRIPAC), National Laboratory of Pattern Recognition (NLPR), Institute of Automation, Chinese Academy of Sciences (CASIA)（中国科学院自动化研究所模式识别国家重点实验室智能感知与计算研究中心）；University of Chinese Academy of Sciences (UCAS)（中国科学院大学）
