請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74453
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 簡韶逸(Shao-Yi Chien) | |
dc.contributor.author | Po-Wei Yen | en |
dc.contributor.author | 嚴柏煒 | zh_TW |
dc.date.accessioned | 2021-06-17T08:36:38Z | - |
dc.date.available | 2019-08-18 | |
dc.date.copyright | 2019-08-18 | |
dc.date.issued | 2019 | |
dc.date.submitted | 2019-08-08 | |
dc.identifier.citation | Z. Hui, X. Wang, and X. Gao, “Fast and accurate single image super- resolution via information distillation network,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018.
J.-B. Huang, A. Singh, and N. Ahuja, “Single image super-resolution from transformed self-exemplars.” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5197–5206, 2015. C. Dong, C. C. Loy, K. He, and X. Tang, “Learning a Deep Convolutional Network for Image Super-Resolution,” Proceedings of European Conference on Computer Vision (ECCV), pp. 184–199, 2014. C. Dong, C. C. Loy, and X. Tang, “Accelerating the Super-Resolution Convolutional Neural Network,” Proceedings of European Conference on Computer Vision (ECCV), 2016. W. Shi, J. Caballero, F. Husza ́r, J. Totz, A. P. Aitken, R. Bishop, D. Rueckert, and Z. Wang, “Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016 J. Kim, J. K. Lee, and K. M. Lee, “Accurate image super-resolution using very deep convolutional networks,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015. C. Ledig, L. Theis, F. Huszar, J. Caballero, A. P. Aitken, A. Tejani, J. Totz, Z. Wang, and W. Shi, “Photo-realistic single image super-resolution using a generative adversarial network,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recongnition,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016. B. Lim, S. Son, H. Kim, S. Nah, and K. M. Lee, “Enhanced deep residual networks for single image super-resolution,” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017 C.-Y. Chih and L.-G. Chen, “Architecture design for real time learning-based super resolution using multi-layer computation,” 2017. J. Yang, J. Wright, T. S. Huang, and Y. Ma, “Image super-resolution via sparse representation,” IEEE Transactions on Image Processing (TIP), 2010. 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.” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 416–423, 2001. J. B. Diederik P. Kingma, “Adam: A method for stochastic optimization,” International Conference on Learning Representations (ICLR), 2014. R. Timofte, E. Agustsson, L. V. Gool, M.-H. Yang, and e. a. Lei Zhang, “Ntire 2017 challenge on single image super-resolution: Methods and results.” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,” Proceedings of IEEE International Conference on Computer Vision (ICCV), pp. 1026–1034, 2015. A. Lavin and S. Gray, “Fast algorithms for convolutional neural networks.” Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016. K.-L. Liu, M.-C. Yang, and S.-Y. Chien, “A real-time fhd learning-based super-resolution system without a frame buffer.” 2017. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74453 | - |
dc.description.abstract | 影像超解析度的目標,是將輸入的低解析度影像,產生出含有銳利邊緣和豐富細節的高解析度影像。由於現今顯示器進步的速度飛快,影像超解析度漸漸成為了一項不可或缺的技術。
影像超解析度是一個困難的非單一解問題,因為一張低解析度影像,可能由多張不同的高解析度影像產生。為了解決這個問題,以深度學習為基底的單一影像超解析度演算法,透過數以萬計來自外部的低解析度對高解析度圖塊配對,學習其中的對應關係,而達到了出色且先進的成果和表現。 在本篇論文中,我們提出了一個改良版深度學習網路,在所有參數量都為12K的網路中,表現是最好的,及基於這個網路的即時超解析度系統架構。此網路首先優化了Information Distillation Network,利用sub-pixel convolution讓整個網路只透過3×3卷積就能實現。高效率的架構則是透過Winograd快速卷積完成。該系統通過Synopsys 32奈米技術的驗證,在200MHz的頻率運作,可達到每秒輸出60張1920×1080 (full HD)影像。有了我們提出的及時超解析度系統,就能將超解析度實作在顯示面板的驅動晶片上,改善顯示系統上的影像問題。 | zh_TW |
dc.description.abstract | Image super-resolution aims to generate good quality high-resolution images with sharp edges and rich details information from input low-resolution images. Due to the rapid growth of the display resolution on edge devices, super-resolution has become an essential technique in many applications. Super-resolution is a well-known ill-posed problem since a single low-resolution image could be generated from more than one high-resolution images. To solve this problem, deep learning-based single image super-resolution methods have achieved outstanding performance and gained state-of-the-art results, by learning an end-to-end mapping function from millions of low-resolution (LR) to high-resolution (HR) image pairs.
We proposed an improved deep-learning architecture that generates highest quality image among other 12K parameters networks as well as a real-time deep- learning based super-resolution system. The proposed architecture inferences a deep-learning model that generate high- resolution images from low-resolution images. This real-time super-resolution system makes it possible to integrate super-resolution operation into display driver carrying out computational photography in display system. The system is verified with Synopsys 32nm educational design kit technology which achieves output resolution of 1920 × 1080 full-HD resolution at 60 frames per second (fps). The proposed system is working at a clock frequency of 200 MHz. | en |
dc.description.provenance | Made available in DSpace on 2021-06-17T08:36:38Z (GMT). No. of bitstreams: 1 ntu-108-R05943125-1.pdf: 16949921 bytes, checksum: dda448f6eee7dedfb11765c42225dbb0 (MD5) Previous issue date: 2019 | en |
dc.description.tableofcontents | Abstract (P.i)
List of Figures (P.vii) List of Tables (P.ix) 1 Introduction (P.1) 1.1 Image Super-Resolution (P.1) 1.2 Methods of Single Image Super-Resolution (P.1) 1.2.1 Non Deep-Learning based Methods (P.2) 1.2.2 Deep-Learning based Methods (P.3) 1.3 Motivation and Design Challenges (P.3) 1.4 Thesis Organization (P.3) 2 Related Works of Deep Learning based Algorithms (P.5) 2.1 Bicubic Interpolation (P.5) 2.2 SRCNN (P.6) 2.3 FSRCNN (P.6) 2.3.1 Deconvolution (P.7) 2.4 ESPCN (P.8) 2.4.1 Sub-pixel Convolution (P.8) 2.5 VDSR (P.8) 2.6 SRResNet (P.9) 2.7 EDSR (P.9) 2.8 IDN (P.9) 2.9 Network Analysis (P.10) 2.9.1 High-Resolution Network (P.10) 2.9.2 Low-Resolution Network (P.11) 2.9.3 Network Decision (P.11) 2.10 Hardware Implementation Related Works (P.11) 3. Baseline Algorithm (P.13) 3.1 Information Distillation Network (P.13) 3.1.1 Enhancement Unit (P.14) 3.1.2 Compression Unit (P.16) 3.2 Training and Implementation Details (P.16) 3.2.1 Training Dataset (P.16) 3.2.2 Network Implementation Details (P.17) 3.3 Results (P.17) 4 Improved Network and Architecture Design (P.21) 4.1 Improved Network (P.21) 4.1.1 Network Details (P.22) 4.1.2 Training Dataset (P.22) 4.1.3 Training (P.23) 4.2 Winograd Fast Convolution (P.23) 4.3 Proposed Architecture (P.25) 4.3.1 Basic Winograd Convolution Unit (P.25) 4.3.2 Block Diagram (P.26) 4.3.3 Computation Strategy (P.26) 4.3.3 Quantization (P.27) 5 Experimental and Implementation Results (P.31) 5.1 Experimental Results (P.31) 5.1.1 Different IDN Configuration (P.31) 5.1.2 Fine-tune Result (P.32) 5.2 Implementation Results (P.32) 6 Conclusion (P.37) Reference (P.39) | |
dc.language.iso | en | |
dc.title | 基於深度學習之超解析度即時系統架構設計 | zh_TW |
dc.title | Efficient Deep Learning based Super-Resolution Real-Time System | en |
dc.type | Thesis | |
dc.date.schoolyear | 107-2 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 楊家驤(Chia-Hsiang Yang),吳安宇(An-Yeu Wu),蔡宗漢(Tsung-Han Tsai) | |
dc.subject.keyword | 超解析度,深度學習, | zh_TW |
dc.subject.keyword | Super-Resolution,Deep Learning, | en |
dc.relation.page | 39 | |
dc.identifier.doi | 10.6342/NTU201902878 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2019-08-11 | |
dc.contributor.author-college | 電機資訊學院 | zh_TW |
dc.contributor.author-dept | 電子工程學研究所 | zh_TW |
顯示於系所單位: | 電子工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-108-1.pdf 目前未授權公開取用 | 16.55 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。