圖像超分辨率是指透過軟體或硬體的方式將圖片從低解析度恢復至高解析度，且恢復的過程中需保留圖像重要細節和相關訊息，以確保恢復後的高解析度圖像不會嚴重失真。近年來基於深度卷積神經網路的快速發展，使得圖像超分辨率有了非常明顯的進步。本文選擇了近代深度卷積神經網路中的殘差密集網路作為基準，並修改殘差密集網路的核心架構為多重路徑，透過多重路徑的學習使得殘差密集網路可以學到更細緻的特徵，藉此改進殘差密集網路的恢復效果以及參數量，我們稱此網路架構為多重殘差密集網路。本文會設計多種的路徑架構，並在幾個常用測試數據集以及不同的退化模型上進行實驗，來比較不同圖像超分辨率的模型性能，以此證明多重殘差密集網路相較於殘差密集網路有更好的性能。

Image super-resolution refers to the restoration of images from low-resolution to high-resolution through software or hardware. To ensure the quality of the generated high-resolution image, it is necessary to preserve the important details and information of the image. In recent years, deep convolutional neural network models have shown superior performance in image super-resolution.

In this paper, we modify the residual dense network architecture by adding N paths to enhance the learning of local features, and named N-path residual dense network. The local features can learn more information to improve the recovery effect and reduce the number of parameters of the residual dense network. This study designs different N-path residual dense network architectures, and use different datasets and degradation models to experiment. The experiment results suggest that the proposed N-path residual dense network perform superior than the base line residual dense network.

[1] Z. Lei and W. Xiaolin, "An edge-guided image interpolation algorithm via
directional filtering and data fusion," IEEE Transactions on Image Processing,
vol. 15, no. 8, pp. 2226-2238, doi: 10.1109/TIP.2006.877407, 2006.
[2] R. Zeyde, M. Elad, and M. Protter, "On Single Image Scale-Up Using SparseRepresentations," in Curves and Surfaces, Berlin, Heidelberg, J.-D. Boissonnat
et al., Eds., Springer Berlin Heidelberg, pp. 711-730, 2012// 2012.
[3] Z. Li, J. Yang, Z. Liu, X. Yang, G. Jeon, and W. Wu, "Feedback Network for
Image Super-Resolution," 2019 IEEE/CVF Conference on Computer Vision
and Pattern Recognition (CVPR), pp. 3862-3871, 2019.
[4] Q. Li, Z. Li, L. Lu, G. Jeon, K. Liu, and X. Yang, "Gated Multiple Feedback
Network for Image Super-Resolution," in BMVC, 2019.
[5] X. Hu, H. Mu, X. Zhang, Z. Wang, T. Tan, and J. Sun, "Meta-SR: A
Magnification-Arbitrary Network for Super-Resolution," 2019 IEEE/CVF
Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1575-
1584, 2019.
[6] Y. Zhang, Y. Tian, Y. Kong, B. Zhong, and Y. Fu, "Residual dense network for
image super-resolution," in Proceedings of the IEEE conference on computer
vision and pattern recognition, pp. 2472-2481, 2018.
[7] K. Zhang, W. Zuo, and L. Zhang, "Learning a Single Convolutional SuperResolution Network for Multiple Degradations," 2018 IEEE/CVF Conference
on Computer Vision and Pattern Recognition, pp. 3262-3271, 2018.
[8] M. Haris, G. Shakhnarovich, and N. Ukita, "Deep Back-Projection Networks
for Super-Resolution," 2018 IEEE/CVF Conference on Computer Vision and
Pattern Recognition, pp. 1664-1673, 2018.
[9] W.-S. Lai, J.-B. Huang, N. Ahuja, and M.-H. Yang, "Deep Laplacian Pyramid
Networks for Fast and Accurate Super-Resolution," 2017 IEEE Conference on
Computer Vision and Pattern Recognition (CVPR), pp. 5835-5843, 2017.
[10] Y. Tai, J. Yang, X. Liu, and C. Xu, "MemNet: A Persistent Memory Network
for Image Restoration," 2017 IEEE International Conference on Computer
Vision (ICCV), pp. 4549-4557, 2017.
[11] B. Lim, S. Son, H. Kim, S. Nah, and K. M. Lee, "Enhanced Deep Residual
Networks for Single Image Super-Resolution," 2017 IEEE Conference on
Computer Vision and Pattern Recognition Workshops (CVPRW), pp. 1132-
1140, 2017.
[12] T. Tong, G. Li, X. Liu, and Q. Gao, "Image Super-Resolution Using Dense Skip
Connections," in 2017 IEEE International Conference on Computer Vision
(ICCV), pp. 4809-4817, doi: 10.1109/ICCV.2017.514, 22-29 Oct. 2017 2017.
[13] R. Timofte et al., "NTIRE 2017 Challenge on Single Image Super-Resolution:
Methods and Results," in 2017 IEEE Conference on Computer Vision and
Pattern Recognition Workshops (CVPRW),pp. 1110-1121, doi:
10.1109/CVPRW.2017.149, 21-26 July 2017 2017.
[14] C. Dong, C. C. Loy, K. He, and X. Tang, "Image Super-Resolution Using Deep
Convolutional Networks," IEEE Transactions on Pattern Analysis and Machine
Intelligence, vol. 38, pp. 295-307, 2016.
[15] C. Dong, C. C. Loy, and X. Tang, "Accelerating the Super-Resolution
Convolutional Neural Network," in ECCV, 2016.
[16] J. Kim, J. K. Lee, and K. M. Lee, "Accurate Image Super-Resolution Using
Very Deep Convolutional Networks," 2016 IEEE Conference on Computer
Vision and Pattern Recognition (CVPR), pp. 1646-1654, 2016.
[17] D. Martin, C. Fowlkes, D. Tal, and J. Malik, A Database of Human Segmented
Natural Images and its Application to Evaluating Segmentation Algorithms and
Measuring Ecological Statistics. 416-423 vol.2, pp, 2001.
[18] J. B. Huang, A. Singh, and N. Ahuja, "Single image super-resolution from
transformed self-exemplars," in 2015 IEEE Conference on Computer Vision
and Pattern Recognition (CVPR), pp. 5197-5206, doi:
10.1109/CVPR.2015.7299156. , 7-12 June 2015 2015.
[19] R. J. Wang, X. Li, S. Ao, and C. X. Ling, "Pelee: A Real-Time Object Detection
System on Mobile Devices," in NeurIPS, 2018.
[20] Z. Wang, D. Liu, J. Yang, W. Han, and T. Huang, "Deep Networks for Image
Super-Resolution with Sparse Prior," in 2015 IEEE International Conference
on Computer Vision (ICCV), pp. 370-378, doi: 10.1109/ICCV.2015.50, 7-13
Dec. 2015 2015.
[21] M. Bevilacqua, A. Roumy, C. Guillemot, and M.-L. Alberi-Morel, "LowComplexity Single Image Super-Resolution Based on Nonnegative Neighbor
Embedding," doi: 10.5244/C.26.135, 09/01 2012.
[22] Y. Matsui et al., "Sketch-based manga retrieval using manga109 dataset,"
Multimedia Tools and Applications, vol. 76, pp. 21811-21838, 2016.
[23] R. Timofte, R. Rothe, and L. V. Gool, "Seven Ways to Improve Example-Based
Single Image Super Resolution," 2016 IEEE Conference on Computer Vision
and Pattern Recognition (CVPR), pp. 1865-1873, 2016.