Video Super-resolution via Hierarchical Feature Reuse

被引：0

作者：

Zhou, Yuan ^{[1
]}

Wang, Ming-Fei ^{[1
]}

Du, Xiao-Ting ^{[1
]}

Chen, Yan-Fang ^{[1
]}

机构：

[1] School of Electrical and Information Engineering, Tianjin University, Tianjin

来源：

Zidonghua Xuebao/Acta Automatica Sinica | 2024年 / 50卷 / 09期

基金：

中国国家自然科学基金;

关键词：

convolutional neural network (CNN); feature fusion; Hierarchical feature reuse; video super-resolution;

D O I：

10.16383/j.aas.c210095

中图分类号：

学科分类号：

摘要：

The performance improvement of current deep convolution neural network methods in video super-resolution task is slightly lower than that in image super-resolution task, partly because they do not make full use of some key inter-frame information in hierarchical structure features. In this paper, we propose hierarchical feature reuse network (HFRNet) to solve the problem mentioned above. The network retains the low-frequency content of the motion compensation frame, and use dense hierarchical feature block (DHFB) to adaptively fuse the features of each residual block within it, then long-term feature reuse is proposed to fuse the features between multiple dense hierarchical feature block, so as to promote the recovery of high-frequency detail information. Experimental results show that the proposed method is superior to the current method in both quantitative and qualitative metrics. © 2024 Science Press. All rights reserved.

引用

页码：1736 / 1746

页数：10

共 41 条

[1]

Liu C, Sun D., On Bayesian adaptive video super resolution, IEEE Transactions on Pattern Analysis and Machine Intelligence, 36, 2, pp. 346-360, (2014)

[2]

Shahar O, Faktor A, Irani M., Space-time super-resolution from a single video, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3353-3360, (2011)

[3]

Zhou Y, Wang Y, Zhang Y, Du X, Liu H, Li C., Manifold learning based super resolution for mixed-resolution multi-view video in visual internet of things, Proceedings of the International Conference on Artificial Intelligence for Communications and Networks, pp. 486-495, (2019)

[4]

Caballero J, Ledig C, Aitken A, Acosta A, Totz J, Wang Z, Et al., Real-time video super-resolution with spatio-temporal networks and motion compensation, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2848-2857, (2017)

[5]

Tao X, Gao H, Liao R, Wang J, Jia J., Detail-revealing deep video super-resolution, Proceedings of the IEEE International Conference on Computer Vision, pp. 4482-4490, (2017)

[6]

Kappeler A, Yoo S, Dai Q, Katsaggelos A K., Video super-resolution with convolutional neural networks, IEEE Transactions on Computational Imaging, 2, 2, pp. 109-122, (2016)

[7]

Li D, Wang Z., Video super resolution via motion compensation and deep residual learning, IEEE Transactions on Computational Imaging, 3, 4, pp. 749-762, (2017)

[8]

Zhou Y, Zhang Y, Xie X, Kung S-Y., Image super-resolution based on dense convolutional auto-encoder blocks, Neurocomputing, 423, 1, pp. 98-109, (2021)

[9]

Li Jin-Xin, Huang Zhi-Yong, Li Wen-Bin, Zhou Deng-Wen, Image super-resolution based on multi hierarchical features fusion network, Acta Automatica Sinica, 49, 1, pp. 161-171, (2023)

[10]

Zhang Yi-Feng, Liu Yuan, Jiang Cheng, Cheng Xu, A curriculum learning approach for single image super resolution, Acta Automatica Sinica, 46, 2, pp. 274-282, (2020)

← 1 2 3 4 5 →