Video Super-Resolution with Recurrent Structure-Detail Network
This work addresses the problem of efficient and effective video super-resolution for applications like video enhancement, with incremental improvements in recurrent-based methods.
The paper tackles video super-resolution by proposing a recurrent method that divides input into structure and detail components, using a recurrent unit with two-stream blocks and a hidden state adaptation module to enhance robustness, achieving superior performance on benchmark datasets compared to state-of-the-art methods.
Most video super-resolution methods super-resolve a single reference frame with the help of neighboring frames in a temporal sliding window. They are less efficient compared to the recurrent-based methods. In this work, we propose a novel recurrent video super-resolution method which is both effective and efficient in exploiting previous frames to super-resolve the current frame. It divides the input into structure and detail components which are fed to a recurrent unit composed of several proposed two-stream structure-detail blocks. In addition, a hidden state adaptation module that allows the current frame to selectively use information from hidden state is introduced to enhance its robustness to appearance change and error accumulation. Extensive ablation study validate the effectiveness of the proposed modules. Experiments on several benchmark datasets demonstrate the superior performance of the proposed method compared to state-of-the-art methods on video super-resolution.