

浏览全部资源
扫码关注微信
1.合肥工业大学 机械工程学院, 安徽 合肥 230002
2.东北大学 机械工程与自动化学院, 辽宁 沈阳 110167
Received:28 December 2025,
Revised:2026-02-02,
Published:05 March 2026
移动端阅览
REN Pengyang, PANG Kai. Lightweight video super-resolution reconstruction based on OFCA-Transformer[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(3): 440-451.
REN Pengyang, PANG Kai. Lightweight video super-resolution reconstruction based on OFCA-Transformer[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(3): 440-451. DOI: 10.37188/CJLCD.2025-0256. CSTR: 32172.14.CJLCD.2025-0256.
针对现有视频超分辨率方法在复杂运动场景下存在的帧间对齐不准确、时序信息利用不充分,以及传统注意力机制计算复杂度高等问题,本文提出一种融合光流引导交叉注意力的视频超分辨率网络(OFCA-Transformer)。首先,设计一个轻量级的多尺度光流估计模块,生成多粒度运动信息;其次,创新性引入光流引导的交叉注意力机制,以光流预测位置为中心建立局部注意力窗口,实现显式几何先验与隐式内容感知的深度融合,在提升对齐精度的同时显著降低计算复杂度;最后,构建分层特征聚合模块,在Transformer架构内实现更有效的时空特征融合。在放大因子分别为×2、×3、×4时,将本文的研究方法与其他方法在3个公开数据集进行对比。结果表明,OFCA-Transformer在多个数据集上的PSNR值与其他先进方法相比仅差0.16 dB,而模型参数量降低82.8%,有效地提高了计算效率。此外,本文所提的研究方法在复杂运动场景下表现出更精确的细节恢复和更好的时间一致性,客观上在各个放大因子下均取得较好的定量指标。
To address the limitations of existing video super-resolution methods in complex motion scenes-including inaccurate frame-to-frame alignment, insufficient utilization of temporal information, and high computational complexity of traditional attention mechanisms, this paper proposes an optical flow-guided cross-attention video super-resolution network (OFCA-Transformer). First, a lightweight multi-scale optical flow estimation module is designed to generate multi-granularity motion information. Second, we innovatively introduce a flow-guided cross-attention mechanism. By establishing local attention windows centered on flow-predicted positions, we achieve an explicit fusion of geometric priors with implicit content awareness. This approach significantly enhances alignment accuracy while substantially reducing computational complexity. Additionally, we construct a hierarchical feature aggregation module to enable more efficient spatio-temporal feature fusion within the Transformer architecture. Our method was evaluated against other approaches on three public datasets at magnification factors of ×2, ×3, and ×4. The results demonstrate that OFCA-Transformer achieves PSNR values only 0.16 dB lower than the state-of-the-art methods across multiple datasets, while reducing model parameters by 82.8%, effectively improving computational efficiency. Furthermore, the proposed method exhibits more precise detail recovery and better temporal consistency in complex motion scenes, objectively achieving superior quantitative metrics across all magnification factors.
唐麒 , 赵耀 , 刘美琴 , 等 . 基于深度学习的视频超分辨率重建算法进展 [J]. 自动化学报 , 2025 , 51 ( 7 ): 1480 - 1524 . doi: 10.16383/j.aas.c240235 http://dx.doi.org/10.16383/j.aas.c240235
TANG Q , ZHAO Y , LIU M Q , et al . A review of video super-resolution algorithms based on deep learning [J]. Acta Automatica Sinica , 2025 , 51 ( 7 ): 1480 - 1524 . (in Chinese) . doi: 10.16383/j.aas.c240235 http://dx.doi.org/10.16383/j.aas.c240235
江俊君 , 程豪 , 李震宇 , 等 . 深度学习视频超分辨率技术综述 [J]. 中国图象图形学报 , 2023 , 28 ( 7 ): 1927 - 1964 . doi: 10.11834/jig.220130 http://dx.doi.org/10.11834/jig.220130
JIANG J J , CHENG H , LI Z Y , et al . Deep learning based video-related super-resolution technique: a survey [J]. Journal of Image and Graphics , 2023 , 28 ( 7 ): 1927 - 1964 . (in Chinese) . doi: 10.11834/jig.220130 http://dx.doi.org/10.11834/jig.220130
WAN Z Y , ZHANG B , CHEN D D , et al . Bringing old films back to life [C]// Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . New Orleans, USA : IEEE , 2022 : 17673 - 17682 . doi: 10.1109/cvpr52688.2022.01717 http://dx.doi.org/10.1109/cvpr52688.2022.01717
LI G , JI J , QIN M H , et al . Towards high-quality and efficient video super-resolution via spatial-temporal data overfitting [C]// Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Vancouver, Canada : IEEE , 2023 : 10259 - 10269 . doi: 10.1109/cvpr52729.2023.00989 http://dx.doi.org/10.1109/cvpr52729.2023.00989
KHAN M A , JAVED K , KHAN S A , et al . Human action recognition using fusion of multiview and deep features: an application to video surveillance [J]. Multimedia Tools and Applications , 2024 , 83 ( 5 ): 14885 - 14911 .
LUO L G , YI B S , WANG Z Y , et al . Efficient lightweight network for video super-resolution [J]. Neural Computing and Applications , 2024 , 36 ( 2 ): 883 - 896 . doi: 10.1007/s00521-023-09065-z http://dx.doi.org/10.1007/s00521-023-09065-z
SUN X , LONG X , HE D L , et al . VSRNet: end-to-end video segment retrieval with text query [J]. Pattern Recognition , 2021 , 119 : 108027 . doi: 10.1016/j.patcog.2021.108027 http://dx.doi.org/10.1016/j.patcog.2021.108027
CHAN K C K , WANG X T , YU K , et al . BasicVSR: the search for essential components in video super-resolution and beyond [C]// Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Nashville, USA : IEEE , 2021 : 4947 - 4956 . doi: 10.1109/cvpr46437.2021.00491 http://dx.doi.org/10.1109/cvpr46437.2021.00491
CHAN K C K , ZHOU S C , XU X Y , et al . BasicVSR++: IMPROVING video super-resolution with enhanced propagation and alignment [C]// Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition . New Orleans : IEEE , 2022 : 5972 - 5981 . doi: 10.1109/cvpr52688.2022.00588 http://dx.doi.org/10.1109/cvpr52688.2022.00588
CAO J Z , LI Y W , ZHANG K , et al . Video super-resolution transformer [J/OL]. arXiv , 2021 : 2106 . 06847 .
WANG X , WANG H , ZHANG M L , et al . Combining optical flow and Swin Transformer for space-time video super-resolution [J]. Engineering Applications of Artificial Intelligence , 2024 , 137 : 109227 . doi: 10.1016/j.engappai.2024.109227 http://dx.doi.org/10.1016/j.engappai.2024.109227
夏振平 , 陈豪 , 张宇宁 , 等 . 基于混合时空卷积的轻量级视频超分辨率重建 [J]. 光学 精密工程 , 2024 , 32 ( 16 ): 2564 - 2576 . doi: 10.37188/OPE.20243216.2564 http://dx.doi.org/10.37188/OPE.20243216.2564
XIA Z P , CHEN H , ZHANG Y N , et al . Lightweight video super-resolution based on hybrid spatio-temporal convolution [J]. Optics and Precision Engineering , 2024 , 32 ( 16 ): 2564 - 2576 . (in Chinese) . doi: 10.37188/OPE.20243216.2564 http://dx.doi.org/10.37188/OPE.20243216.2564
林坚普 , 吴镇城 , 王崑赋 , 等 . 级联残差优化Transformer网络的图像超分辨率重建 [J]. 光学 精密工程 , 2024 , 32 ( 12 ): 1902 - 1914 . doi: 10.37188/OPE.20243212.1902 http://dx.doi.org/10.37188/OPE.20243212.1902
LIN J P , WU Z C , WANG K F , et al . Cascade residual-optimized image super-resolution reconstruction in Transformer network [J]. Optics and Precision Engineering , 2024 , 32 ( 12 ): 1902 - 1914 . (in Chinese) . doi: 10.37188/OPE.20243212.1902 http://dx.doi.org/10.37188/OPE.20243212.1902
XUE T F , CHEN B A , WU J J , et al . Video enhancement with task-oriented flow [J]. International Journal of Computer Vision , 2019 , 127 ( 8 ): 1106 - 1125 . doi: 10.1007/s11263-018-01144-2 http://dx.doi.org/10.1007/s11263-018-01144-2
陈清江 , 陈鹏民 . 多维度聚合Transformer的图像超分辨率重建 [J]. 光学 精密工程 , 2025 , 33 ( 12 ): 1955 - 1970 . doi: 10.37188/ope.20253312.1955 http://dx.doi.org/10.37188/ope.20253312.1955
CHEN Q J , CHEN P M . MDAT: multi-dimensional aggregation Transformer for image super-resolution reconstruction [J]. Optics and Precision Engineering , 2025 , 33 ( 12 ): 1955 - 1970 . (in Chinese) . doi: 10.37188/ope.20253312.1955 http://dx.doi.org/10.37188/ope.20253312.1955
JIN B Y , HOU Y W , XI P . Bridging microscale and macroscale light-field image reconstruction using ‘real-time and universal network’ [J]. Light: Advanced Manufacturing , 2025 , 6 ( 3 ): 399 - 401 . doi: 10.37188/lam.2025.060 http://dx.doi.org/10.37188/lam.2025.060
贺兴 , 王磊 , 张鹏超 , 等 . 基于多维注意力网络的图像超分辨率重建 [J]. 液晶与显示 , 2025 , 40 ( 7 ): 1056 - 1066 . doi: 10.37188/cjlcd.2025-0058 http://dx.doi.org/10.37188/cjlcd.2025-0058
HE X , WANG L , ZHANG P C , et al . Image super-resolution reconstruction based on multidimensional attention network [J]. Chinese Journal of Liquid Crystals and Displays , 2025 , 40 ( 7 ): 1056 - 1066 . (in Chinese) . doi: 10.37188/cjlcd.2025-0058 http://dx.doi.org/10.37188/cjlcd.2025-0058
阎刚 , 宋子怡 , 耿树泽 . 基于全方位状态空间模型的轻量化图像超分辨率重建 [J]. 液晶与显示 , 2025 , 40 ( 4 ): 642 - 654 . doi: 10.37188/cjlcd.2024-0246 http://dx.doi.org/10.37188/cjlcd.2024-0246
YAN G , SANG Z Y , GENG S Z . PMambaIR: panoramic vision state space model for lightweight image super-resolution [J]. Chinese Journal of Liquid Crystals and Displays , 2025 , 40 ( 4 ): 642 - 654 . (in Chinese) . doi: 10.37188/cjlcd.2024-0246 http://dx.doi.org/10.37188/cjlcd.2024-0246
YANG M Q , WANG Y H , YANG Y , et al . Multi-scale spatiotemporal feature fusion for super-resolution video reconstruction in dynamic scenes [J]. Engineering Applications of Artificial Intelligence , 2025 , 161 : 112327 . doi: 10.1016/j.engappai.2025.112327 http://dx.doi.org/10.1016/j.engappai.2025.112327
SUN D Q , YANG X D , LIU M Y , et al . PWC-Net: CNNs for optical flow using pyramid, warping, and cost volume [C]// Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Salt Lake City : IEEE , 2018 : 8934 - 8943 . doi: 10.1109/cvpr.2018.00931 http://dx.doi.org/10.1109/cvpr.2018.00931
RANJAN A , BLACK M J . Optical flow estimation using a spatial pyramid network [C]// Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Honolulu : IEEE , 2017 : 2720 - 2729 . doi: 10.1109/cvpr.2017.291 http://dx.doi.org/10.1109/cvpr.2017.291
TIAN Y P , ZHANG Y L , FU Y , et al . TDAN: temporally-deformable alignment network for video super-resolution [C]// Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Seattle : IEEE , 2020 : 3357 - 3366 . doi: 10.1109/cvpr42600.2020.00342 http://dx.doi.org/10.1109/cvpr42600.2020.00342
WANG X L , GIRSHICK R , GUPTA A , et al . Non-local neural networks [C]// Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Salt Lake City : IEEE , 2018 : 7794 - 7803 . doi: 10.1109/cvpr.2018.00813 http://dx.doi.org/10.1109/cvpr.2018.00813
JOHNSON J , ALAHI A , FEI-FEI L . Perceptual losses for real-time style transfer and super-resolution [C]// Proceedings of the 14th European Conference on Computer Vision . Amsterdam : Springer , 2016 : 694 - 711 . doi: 10.1007/978-3-319-46475-6_43 http://dx.doi.org/10.1007/978-3-319-46475-6_43
SIMONYAN K , ZISSERMAN A . Very deep convolutional networks for large-scale image recognition [C]// Proceedings of the 3rd International Conference on Learning Representations . San Diego : OpenReview.net , 2015 : 1 - 14 .
NAH S , BAIK S , HONG S , et al . NTIRE 2019 challenge on video deblurring and super-resolution: dataset and study [C]// Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) . Long Beach : IEEE , 2019 : 1996 - 2005 . doi: 10.1109/cvprw.2019.00251 http://dx.doi.org/10.1109/cvprw.2019.00251
LIU C , SUN D Q . On Bayesian adaptive video super resolution [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence , 2014 , 36 ( 2 ): 346 - 360 . doi: 10.1109/tpami.2013.127 http://dx.doi.org/10.1109/tpami.2013.127
0
Views
880
下载量
0
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621