1.东华理工大学 信息工程学院, 江西 南昌330013
[ "蔡建枫(1997—),男,广东潮州人,硕士研究生,2021年于韩山师范学院获得学士学位,主要从事图像超分辨率重建方面的研究。E-mail:cjfv@qq.com" ]
[ "蒋年德(1971—),男,广西全州人,博士,副教授,2010年于湖南大学获得博士学位,主要从事计算机网络与分布式数据库的应用、数字图像处理与信息融合方面的研究。E-mail:cjnd@163.com" ]
扫 描 看 全 文
蔡建枫, 蒋年德. 基于退化感知的盲超分辨率模型[J]. 液晶与显示, 2023,38(9):1224-1233.
CAI Jian-feng, JIANG Nian-de. Blind super-resolution model based on degradation-aware[J]. Chinese Journal of Liquid Crystals and Displays, 2023,38(9):1224-1233.
蔡建枫, 蒋年德. 基于退化感知的盲超分辨率模型[J]. 液晶与显示, 2023,38(9):1224-1233. DOI: 10.37188/CJLCD.2022-0385.
CAI Jian-feng, JIANG Nian-de. Blind super-resolution model based on degradation-aware[J]. Chinese Journal of Liquid Crystals and Displays, 2023,38(9):1224-1233. DOI: 10.37188/CJLCD.2022-0385.
现有的超分辨率方法假设了从高分辨率图像到低分辨率图像的预定义退化过程,这对于具有复杂退化类型的真实世界图像来说很难成立。针对该问题,本文提出了基于退化感知的盲超分辨率模型。该模型用随机模糊核生成低分辨率图像,用对比学习退化表征。模型生成器由包含多个退化感知块的残差组构成。退化感知块用退化表征和图像特征做交叉注意力计算空间权重图,此外模型还收集残差组输出的层级特征,计算层间注意力来复用层级特征,使得模型更加关注高频细节,模型特征提取能力进一步提高。通过消融实验验证了各模块的有效性。在多个国际公开测试集上,放大倍数为4的平均PSNR提升1.45 dB,SSIM提升0.058。实验结果表明,该模型在盲超分辨率任务上取得了显著的性能,且具有良好的视觉效果。
Existing super-resolution methods assume a predefined degradation process from high-resolution images to low-resolution images, which is difficult to hold for real-world images with complex degradation types. For this problem, a blind super-resolution model based on degradation-aware is proposed. The model generates low-resolution images with random blur kernels and learns degenerate representations with contrasts. The model generator consists of residual groups containing multiple degradation-aware blocks. Degraded perceptual blocks use degraded representations and image features to do cross-attention to calculate spatial weight maps. In addition, the model collects layer-level features from the output of residual groups and calculates inter-layer attention to reuse layer-level features. This enables the model to pay more attention to high-frequency details, and the model feature extraction capability is further improved. The effectiveness of each module is verified by ablation experiments. On multiple international public test sets, the average PSNR with a magnification of 4 is increased by 1.45 dB, and the SSIM is increased by 0.058. Experimental results show that the model achieves significant performance on blind super-resolution tasks with good visual results.
超分辨率盲超分辨率交叉注意力空间注意力特征复用
super resolutionblind super resolutioncross attentionspatial attentionfeature reuse
毕勇,潘鸣奇,张硕,等. 三维点云数据超分辨率技术[J]. 中国光学,2022,15(2):210-223. doi: 10.37188/CO.2021-0176http://dx.doi.org/10.37188/CO.2021-0176
BI Y, PAN M Q, ZHANG S, et al. Overview of 3D point cloud super-resolution technology [J]. Chinese Optics, 2022, 15(2): 210. (in Chinese). doi: 10.37188/CO.2021-0176http://dx.doi.org/10.37188/CO.2021-0176
DONG C, LOY C C, HE K, et al. Image super-resolution using deep convolutional networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 38(2): 295-307. doi: 10.1109/tpami.2015.2439281http://dx.doi.org/10.1109/tpami.2015.2439281
KIM J, LEE J K, LEE K M. Deeply-recursive convolutional network for image super-resolution [C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA: IEEE, 2016: 1637-1645. doi: 10.1109/cvpr.2016.181http://dx.doi.org/10.1109/cvpr.2016.181
LIM B, SON S, KIM H, et al. Enhanced deep residual networks for single image super-resolution [C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Honolulu, HI, USA: IEEE, 2017: 1132-1140. doi: 10.1109/cvprw.2017.151http://dx.doi.org/10.1109/cvprw.2017.151
DONG C, LOY C C, HE K M, et al. Learning a deep convolutional network for image super-resolution [C].13th European Conference on Computer Vision. Zurich: Springer, 2014: 184-199. doi: 10.1007/978-3-319-10593-2_13http://dx.doi.org/10.1007/978-3-319-10593-2_13
HUI Z, WANG X M, GAO X B. Fast and accurate single image super-resolution via information distillation network [C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018: 723-731. doi: 10.1109/cvpr.2018.00082http://dx.doi.org/10.1109/cvpr.2018.00082
QIU Y J, WANG R X, TAO D P, et al. Embedded block residual network: a recursive restoration model for single-image super-resolution [C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Seoul, Korea (South): IEEE, 2019: 4179-4188. doi: 10.1109/iccv.2019.00428http://dx.doi.org/10.1109/iccv.2019.00428
LIU J, ZHANG W J, TANG Y T, et al. Residual feature aggregation network for image super-resolution [C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, WA, USA: IEEE, 2020: 2356-2365. doi: 10.1109/cvpr42600.2020.00243http://dx.doi.org/10.1109/cvpr42600.2020.00243
BELL-KLIGLER S, SHOCHER A, IRANI M. Blind super-resolution kernel estimation using an internal-GAN [C]//Proceedings of the 33rd International Conference on Neural Information Processing Systems. Vancouver: Curran Associates Inc., 2019: 26.
ZHU F, NING Y, CHEN X C, et al. On removing potential redundant constraints for SVOR learning [J]. Applied Soft Computing, 2021, 102: 106941. doi: 10.1016/j.asoc.2020.106941http://dx.doi.org/10.1016/j.asoc.2020.106941
EFRAT N, GLASNER D, APARTSIN A, et al. Accurate blur models vs. image priors in single image super-resolution [C]//Proceedings of the 2013 IEEE International Conference on Computer Vision. Sydney, NSW, Australia: IEEE, 2013: 2832-2839. doi: 10.1109/iccv.2013.352http://dx.doi.org/10.1109/iccv.2013.352
YANG C Y, MA C, YANG M H. Single-image super-resolution: a benchmark [C]. 13th European Conference on Computer Vision. Zurich: Springer, 2014: 372-386. doi: 10.1109/cvpr.2019.00170http://dx.doi.org/10.1109/cvpr.2019.00170
ZHANG K, ZUO W M, ZHANG L. Learning a single convolutional super-resolution network for multiple degradations [C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018: 3262-3271. doi: 10.1109/cvpr.2018.00344http://dx.doi.org/10.1109/cvpr.2018.00344
SHOCHER A, COHEN N, IRANI M. Zero-shot super-resolution using deep internal learning [C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018: 3118-3126. doi: 10.1109/cvpr.2018.00329http://dx.doi.org/10.1109/cvpr.2018.00329
ZHANG K, VAN GOOL L, TIMOFTE R. Deep unfolding network for image super-resolution [C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, WA, USA: IEEE, 2020: 3214-3223. doi: 10.1109/cvpr42600.2020.00328http://dx.doi.org/10.1109/cvpr42600.2020.00328
ZHANG K, ZUO W, ZHANG L. Deep plug-and-play super-resolution for arbitrary blur kernels [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, IEEE, 2019: 1671-1681. doi: 10.1109/cvpr.2019.00177http://dx.doi.org/10.1109/cvpr.2019.00177
GU J J, LU H N, ZUO W M, et al. Blind super-resolution with iterative kernel correction [C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach, CA, USA: IEEE, 2019: 1604-1613. doi: 10.1109/cvpr.2019.00170http://dx.doi.org/10.1109/cvpr.2019.00170
WANG L G, WANG Y Q, DONG X Q, et al. Unsupervised degradation representation learning for blind super-resolution [C]//Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville, TN, USA: IEEE, 2021: 10576-10585. doi: 10.1109/cvpr46437.2021.01044http://dx.doi.org/10.1109/cvpr46437.2021.01044
ELAD M, FEUER A. Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images [J]. IEEE Transactions on Image Processing, 1997, 6(12): 1646-1658. doi: 10.1109/83.650118http://dx.doi.org/10.1109/83.650118
HE K M, FAN H Q, WU Y X, et al. Momentum contrast for unsupervised visual representation learning [C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, WA, USA: IEEE, 2020: 9726-9735. doi: 10.1109/cvpr42600.2020.00975http://dx.doi.org/10.1109/cvpr42600.2020.00975
CHEN Y P, DAI X Y, CHEN D D, et al. Mobile-former: bridging mobileNet and transformer [C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans, LA, USA: IEEE, 2022: 5260-5269. doi: 10.1109/cvpr52688.2022.00520http://dx.doi.org/10.1109/cvpr52688.2022.00520
董本志,于明聪,赵鹏. 基于小波域的图像超分辨率重建方法[J]. 液晶与显示,2021,36(2):317-326. doi: 10.37188/CJLCD.2020-0101http://dx.doi.org/10.37188/CJLCD.2020-0101
DONG B Z, YU M C, ZHAO P. Image super-resolution reconstruction based on wavelet domain [J]. Chinese Journal of Liquid Crystals and Displays, 2021, 36(2): 317-326.(in Chinese). doi: 10.37188/CJLCD.2020-0101http://dx.doi.org/10.37188/CJLCD.2020-0101
WANG Z W, LIU D, YANG J C, et al. Deep networks for image super-resolution with sparse prior [C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Santiago, Chile: IEEE, 2015: 370-378. doi: 10.1109/iccv.2015.50http://dx.doi.org/10.1109/iccv.2015.50
LAI W S, HUANG J B, AHUJA N, et al. Deep laplacian pyramid networks for fast and accurate super-resolution [C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, USA: IEEE, 2017: 5835-5843. doi: 10.1109/cvpr.2017.618http://dx.doi.org/10.1109/cvpr.2017.618
SAJJADI M S M, SCHÖLKOPF B, HIRSCH M. EnhanceNet: single image super-resolution through automated texture synthesis [C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Venice, Italy: IEEE, 2017: 4501-4510. doi: 10.1109/iccv.2017.481http://dx.doi.org/10.1109/iccv.2017.481
ZHANG Y L, LI K P, LI K, et al. Image super-resolution using very deep residual channel attention networks [C]//Proceedings of the 15th European Conference on Computer Vision. Munich: Springer, 2018: 294-310. doi: 10.1007/978-3-030-01234-2_18http://dx.doi.org/10.1007/978-3-030-01234-2_18
NIU B, WEN W L, REN W Q, et al. Single image super-resolution via a holistic attention network [C]. 16th European Conference on Computer Vision. Glasgow: Springer, 2020: 191-207. doi: 10.1007/978-3-030-58610-2_12http://dx.doi.org/10.1007/978-3-030-58610-2_12
0
浏览量
154
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构