Gated Bi-Recurrent Separate Network for Video Super-Resolution

Gao, Lihang; Qian, Jiajia; Xu, Fan; Yan, Jiqiang; Yu, Tianqi; Shi, Fanwei; Hu, Jianling

doi:10.1109/wcsp52459.2021.9613243

Cited by 1 publication

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To make full use of the time-domain information of the video, our FGBRSN adopts bi-directional recurrent propagation. Previous studies [11], [17], [20], [27] have pointed out the superiority of the recurrent structure compared with the sliding-window structure. Specifically, our recurrent unit follows the GBRSN [11] with the structure of GRU [33].…”

Section: Recurrent Propagation Modulementioning

confidence: 99%

“…Previous studies [11], [17], [20], [27] have pointed out the superiority of the recurrent structure compared with the sliding-window structure. Specifically, our recurrent unit follows the GBRSN [11] with the structure of GRU [33]. The introduction of the GRU in the recurrent unit can deal with the problems of short-time memory and error features due to the gating structure.…”

Section: Recurrent Propagation Modulementioning

confidence: 99%

“…Most of the existing methods implement explicit alignment based on the optical flow [15]- [17] or implicit alignment [8], [18], [19] via deformable convolution. Moreover, some other methods [11], [20], [22] adopt a recurrent structure to better utilize the temporal information of video sequences. Flowbased methods estimate the optical flow and warp neighboring frames to align with the reference one.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

FGBRSN: Flow-Guided Gated Bi-Directional Recurrent Separated Network for Video Super-Resolution

Xue,

Gao,

et al. 2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

Video Super-Resolution (VSR) is the task of reconstructing high-resolution (HR) video sequences from low-resolution (LR) video sequences. Apart from spatial information of reference frames, temporal information of neighboring frames is also important for reconstruction. Current VSR methods usually take advantage of the temporal information through optical flow estimation and compensation. However, optical flow estimation is often inaccurate and difficult, which may result in artifacts or blurring in the reconstructed frames. In this paper, we propose a Novel Flow-guided Deformable Alignment Module (NFDAM) for frame alignment. In this module, a lightweight optical network FNet is designed to estimate optical flow as a coarse offset, which then guides the deformable convolution at the feature and image level. On this basis, we propose a Flow-guided Gated Bidirectional Recurrent Separated Network (FGBRSN) for VSR, in which a gated recurrent structure is designed to leverage long-term information and a multilevel residual fusion approach is used in upsampling module. Our experiments on public datasets show that the proposed method improves both quantitative evaluation and visual effects compared with the existing methods.

show abstract