Abstract To expand the perception field with the filter parameters unchanged, dilated convolution is introduced into very deep convolutional networks super-resolution model. Firstly, the perception field of the dilated convolution block with different expansion coefficients is analyzed and a better combination structure is selected as the dilated convolution block. Then, the deep convolution network is constructed by stacking convolution blocks and adding residual connection. Experiment shows that the reconstruction effect can be improved by the constructed network for the larger scaling factors of Set5 dataset. Besides, there are obvious visual advantages.
Fund:Supported by National Natural Science Foundation of China(No.61672477), Natural Science Foundation of Zhejiang Province(No.LY18F020018)
Corresponding Authors:
CAO Feilong, Ph.D., professor. His research interests include deep learning and image processing.
About author:: (ZHANG Zhuolin, master student. His research interests include deep learning and image processing.) (ZHAO Jianwei, Ph.D., professor. Her research interests include deep learning and image processing.)
ZHANG Zhuolin,ZHAO Jianwei,CAO Feilong. Building Deep Neural Networks with Dilated Convolutions to Reconstruct High-Resolution Image[J]. , 2019, 32(3): 259-267.
[1] HUANG J B, SINGH A, AHUJA N. Single Image Super-Resolution from Transformed Self-exemplars // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2015, VI: 5197-5206. [2] SHOCHER A, COHEN N, IRANI M. "Zero-Shot" Super-Resolution Using Deep Internal Learning[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1712.06087.pdf. [3] YANG J C, WRIGHT J, HUANG T S, et al. Image Super-Resolution via Sparse Representation. IEEE Transactions on Image Processing, 2010, 19(11): 2861-2873. [4] TIMOFTE R, DE SMET V, VAN GOOL L. Anchored Neighborhood Regression for Fast Example-Based Super-Resolution // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2013: 1920-1927. [5] TIMOFTE R, DE SMET V, VAN GOOL V. A+: Adjusted Anchored Neighborhood Regression for Fast Super-Resolution // Proc of the Asian Conference on Computer Vision. Berlin, Germany: Springer, 2014: 111-126. [6] ZHANG Y B, ZHANG Y L, ZHANG J, et al. Collaborative Representation Cascade for Single-Image Super-Resolution. IEEE Transaction on Systems, Man, and Cybernetics(Systems), 2017. DOI: 10.1109/TSMC.2017.270548. [7] TIMOFTE R, ROTHE R, VAN GOOL L. Seven Ways to Improve Example-Based Single Image Super Resolution // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2016: 1865-1873. [8] HUANG J J, SIU W C. Learning Hierarchical Decision Trees for Single Image Super-Resolution. IEEE Transaction on Circuits and Systems for Video Technology, 2017, 27(5): 937-950. [9] DONG C, LOY C C, HE K M, et al. Learning a Deep Convolutio-nal Network for Image Super-Resolution // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2014: 184-199. [10] WANG Z W, LIU D, YANG J C, et al. Deep Networks for Image Super-Resolution with Sparse Prior // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2015: 370-378. [11] KIM J, LEE J K, LEE K M. Accurate Image Super-Resolution Using Very Deep Convolutional Networks[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1511.04587.pdf. [12] KIM J, LEE J K, LEE K M. Deeply-Recursive Convolutional Network for Image Super-Resolution[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1511.04491.pdf. [13] BOYD S, PARIKH N, CHU E, et al. Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers. Foundation & Trends in Machine Learning, 2011, 3(1): 1-122. [14] YU F, KOLTUN V. Multi-scale Context Aggregation by Dilated Convolutions[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1511.07122.pdf. [15] LONG J, SHELHAMER E, DARRELL T. Fully Convolutional Networks for Semantic Segmentation[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1411.4038.pdf.
[16] BADRINARAYANAN V, HANDA A, CIPOLLA R. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Robust Semantic Pixel-Wise Labelling[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1505.07293.pdf. [17] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet Cla-ssification with Deep Convolutional Neural Networks. Communications of the ACM, 2017, 60(6): 84-90. [18] KINGMA D K, BA J L. Adam: A Method for Stochastic Optimization[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1412.6980.pdf. [19] HE K M, ZHANG X Y, REN S Q, et al. Deep Residual Learning for Image Recognition // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2016: 770-778. [20] SAXE A M, MCCLELLAND J L, GANGULI S. Exact Solutions to the Nonlinear Dynamics of Learning in Deep Linear Neural Networks[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1312.6120.pdf. [21] KEYS R. Cubic Convolution Interpolation for Digital Image Processing. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1981, 29(6): 1153-1160. [22] JIA Y Q, SHELHAMER E, DONAHUE J, et al. Caffe: Convolutional Architecture for Fast Feature Embedding[C/OL]. [2018-10-22]. https://arxiv.org/pdf/1408.5093.pdf. [23] HE K M, ZHANG X Y, REN S Q, et al. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2015: 1026-1034.
2019, Vol. 32 
