基於自適性稀疏表示之影像顯著度偵測系統超大型積體電路設計

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：20

、訪客IP：3.145.71.161

姓名

林彥宇(Yan-yu Lin) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

基於自適性稀疏表示之影像顯著度偵測系統超大型積體電路設計
(VLSI Architecture Design for Saliency Detection Based on Self-adaptive Sparse Representation)

相關論文

★ Single and Multi-Label Environmental Sound Recognition with Gaussian Process	★ 波束形成與音訊前處理之嵌入式系統實現
★ 語音合成及語者轉換之應用與設計	★ 基於語意之輿情分析系統
★ 高品質口述系統之設計與應用	★ 深度學習及加速強健特徵之CT影像跟骨骨折辨識及偵測
★ 基於風格向量空間之個性化協同過濾服裝推薦系統	★ RetinaNet應用於人臉偵測
★ 金融商品走勢預測	★ 整合深度學習方法預測年齡以及衰老基因之研究
★ 漢語之端到端語音合成研究	★ 基於 ARM 架構上的 ORB-SLAM2 的應用與改進
★ 基於深度學習之指數股票型基金趨勢預測	★ 探討財經新聞與金融趨勢的相關性
★ 基於卷積神經網路的情緒語音分析	★ 運用深度學習方法預測阿茲海默症惡化與腦中風手術存活

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 ( 永不開放)

摘要(中)

本篇論文提出了一個高效能的VLSI架構來實現基於自適性稀疏表示的影像顯著度偵測系統。顯著度偵測是一項建立在人類視覺系統上的重要技術，在本論文中將此系統分為兩個處理階段：特徵表示階段與顯著度測量階段。此二階段在顯著度模型上皆是相當重要的研究課題，如何提出較好的特徵表示法及適當的顯著度測量法更是視覺顯著度模型中的核心問題。在眾多的研究中，稀疏表示已經能夠正確地表示出一個信號的重要部分，所以本研究在特徵表示的選擇上使用稀疏表示法。
在系統流程中，特徵表示階段利用K-SVD演算法來找出資料的稀疏特徵，而在顯著度測量階段利用每個稀疏特徵找出其Background Firing Rate (BFR)，再以Feature Activation Rate (FAR)完成此Bottom-up顯著度偵測系統。
本論文所提出的晶片設計中包含了一個K-SVD模組、一個OMP模組、一個BFR模組和一個FAR模組。晶片的電路設計實現於TSMC 90 nm CMOS Technology，整體的晶片面積約為2.42×2.42 mm2。

摘要(英)

This work proposes an efficient VLSI architecture to perform saliency detection based on sparse representation approach. Saliency detection is a very important technology in the human visual system. Representation and measurement are two important issues for saliency models, and good representation is a critical issue in modelling visual saliency mechanism. Sparse representation has been shown to correctly represent an important part of the signal in a number of studies. This paper utilizes K-SVD algorithm for the feature representation stage, and in saliency measurement stage, background firing rate (BFR) is for each sparse feature and then feature activation rate (FAR) completes the bottom-up saliency detection. The proposed chip comprises a K-SVD module, an OMP module, a BFR module, and a FAR module. The prototype chip is a semi-custom chip that is fabricated using TSMC 90 nm CMOS technology on a die with a size of approximately 2.42x2.42 mm2.

關鍵字(中)

★ 稀疏表示
★ K-SVD
★ 顯著度偵測

關鍵字(英)

論文目次

摘要 ii
Abstract iii
圖目錄 iv
表目錄 vi
章節目次 vii
第一章緒論 1
1.1 前言 1
1.2 研究動機與目的 1
1.3 論文架構 2
第二章顯著度偵測簡介與相關研究 4
2.1 簡介(Introduction) 4
2.2 顯著度偵測相關研究 4
第三章基於自適性稀疏表示偵測演算法 6
3.1 基於自適性稀疏表示偵測演算法流程 6
3.2 特徵表示階段 9
3.3 顯著度測量階段 9

第四章顯著度偵測系統之架構設計 11
4.1 系統架構 11
4.2 K-SVD 12
4.2.1 OMP Unit 17
4.2.2 Matrix Multiplieration Acceleration Unit 22
4.2.3 Omega & Error Unit 25
4.2.4 SVD Unit 26
4.2.5 Selection Unit 31
4.3 BFR Unit 32
4.4 FAR Unit 33
第五章實驗結果 35
5.1 電路環境設置 35
5.2 軟體與硬體之實作及比較 39
5.3 電路的合成與Layout 42
第六章結論及未來研究方向 47
6.1 總結 47
6.2 未來研究方向 47
參考文獻 49

參考文獻

[1] L. Itti, “Automatic Foveation for Video Compression Using aNeurobiological Model of Visual Attention,” IEEE Trans. on Image Processing, vol. 13, no. 10, pp. 1304–1318, Oct. 2004.
[2] A. Treisman and G. Gelade, “A Feature-Integration Theory of Attention,” Cognitive Psychology, vol. 12, no. 1, pp. 97–136, Jan. 1980.
[3] C. Koch and S. Ullman, “Shifts in Selective Visual Attention: Towards the Underlying Neural Circuitry,” Human Neurobiology, vol. 4, pp. 219–227, 1985.
[4] L. Itti, C. Koch, and E. Niebur, “A Model of Saliency-Based Visual Attention for Rapid Scene Analysis,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 20, no. 11, pp. 1254–1259, Nov. 1998.
[5] L. Itti and C. Koch, “Computational Modelling of Visual Attention,” Nature Rev. Neuroscience, vol. 2, no. 3, pp. 194–204, Mar. 2001.
[6] Y. F. Ma and H. J. Zhang, “Contrast-Based Image Attention Analysis by Using Fuzzy Growing,” Proc. 11th ACM Int’l Conf. Multimedia, pp. 374–381, 2003.
[7] D. Walther and C. Koch, “Modeling Attention to Salient Proto-Objects,” Neural Networks, vol. 19, no. 9, pp. 1395–1407, Nov. 2006.
[8] N. D. B. Bruce and J. K.Tsotsos, “Saliency Based on Information Maximization,” Advances in Neural Information Processing Systems, vol. 18, pp. 155–162, Dec. 2005.
[9] J. Harel, C. Koch, and P. Perona, “Graph-Based Visual Saliency,” Advances in Neural Information Processing Systems, vol. 19, pp. 545–552, 2007.
[10] O. L.Meur, P. L.Callet, D. Barba, and D. Thoreau, “A Coherent Computational Approach to Model Bottom-up Visual Attention,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 28, no. 5, pp. 802–817, May 2006.
[11] R. Achanta, S. Hemami, F. Estrada, and S. Susstrunk, “Frequency Tuned Salient Region Detection,” Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 1597–1604, Jun. 2009.
[12] X. Hou and L. Zhang, “Saliency Detection: A Spectral Residual Approach,” Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 1–8, Jun. 2007.
[13] C. Guo, Q. Ma, and L. Zhang, “Spatio-Temporal Saliency Detection Using Phase Spectrum of Quaternion Fourier Transform,” Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 1–8, 2008.
[14] L. Itti, “Models of Bottom-Up and Top-down Visual Attention,” Ph.D. dissertation, Dept. Computat. Neur. Syst., California Inst.,Technol, Pasadena, 2000.
[15] L. Itti, C. Koch, and E. Niebur, “A Model of Saliency-Based Visual Attention for Rapid Scene Analysis,” IEEE Trans. on Pattern Anal. Mach. Intell., vol. 20, no. 11, pp. 1254–1259, Nov. 1998.
[16] R. Achanta and S. Susstrunk, “Saliency Detection for Content-Aware Image Resizing,” Proc.IEEE Int’l Conf. ICIP, pp. 1005–1008, Nov. 2009.
[17] X. Shi, N. D.B. Bruce, and J. K. Tsotsos, “Biologically Motivated Local Contextual Modulation Improves Low-Level Visual Feature Representations,” Lecture Notes in Computer Science, vol. 7324, pp. 79–88, 2012.
[18] S. C. Douglas, M. Gupta, H. Sawada, and S. Makino, “Spatio-Temporal FastICA Algorithms for the Blind Separation of Convolutive Mixtures,” IEEE Trans. on Audio, Speech and Language Processing, Vol. 15, No. 5, pp. 1511–1520, Jul. 2007.
[19] L. M. Ledesma-Carrillo, E. Cabal-Yepez, R. de J Romero-Troncoso, A. Garcia-Perez, R. A. Osornio-Rios and T. D. Carozzi, “Reconfigurable FPGA-Based Unit for Singular Value Decomposition of Large m x n Matrices,” Proc. IEEE Int’l Conf. on ReConFig, pp. 345–350, 2011.
[20] Y. C. Chen, “VLSI Architecture Design for Blind Source Separation Based on Infomax and Time-Frequency Masking,” MS Thesis, NCU, July 2012.
[21] S. T. Huang, “VLSI Architecture Design for Blind Source Separation Based on Time-Frequency Masking and Dictionary Training with Orthogonal Matching Pursuit,” MS Thesis, NCU, July 2013.
[22] J. R. Cavallaro, M. P. Keleher, R. H. Price, and G. S. Thomas “VLSI Implementation of A CORDIC SVD Processor,” Proc. 8th Biennial University/Government/Industry Microelectronics Symp., pp.256–260, Jun.1989.
[23] N. D. Hemkumar, K. Kota, and J. R. Cavallaro “CAPE-VLSI Implementation of aSystolic Processor Array: Architecture, Design and Testing,” Proc. 9th IEEE Biennial University/Government/Industry Microelectronics Symp., pp.64–69, Jun.1991.
[24] Z. Liu, K. Dickson, and J. V. McCanny, “CORDIC Based Application Specific Instruction Set Processor for QRD/SVD,” Proc. 37th Asilomar Conf. on Signals, Systems & Computers, Capacity of Multi-Antenna Gaussian Channels vol.2, pp.1456–1460, Nov. 2003.
[25] Y. Wang, K. Cunningham, P. Nagvajara, and J. Johnson “Singular Value Decomposition Hardware for MIMO: State of the Art and Custom Design,” Proc. IEEE Conf. on ReConFig, pp.400–405, Dec. 2010.
[26] I. E.Telatar. “Capacity of Multi-Antenna Gaussian Channels,” Europ. Trans. on Telecommun, Capacity of Multi-Antenna Gaussian Channels vol. 10, no. 6, pp.585–596, 1999.
[27] H. Li, Y. H. Shen, and S. Z. Chen, “A Robust On-line Blind Separation Algorithm with Dynamic Source Number Based on Neural Network,”Proc. Int’l Conf. onNetworks Security Wireless Communications and Trusted Computing (NSWCTC), vol. 1, pp. 99–102,Apr. 2010
[28] J. R. Cavallaro, M. P. Keleher, R. H. Price, and G. S. Thomas, “VLSI Implementation of a CORDIC SVD Processor,” Proc. 8th Biennial University/Government/Industry Microelectronics Symp.,pp. 256–260, Jun. 1989.
[29] N. D. Hemkumar, K. Kota, and J. R. Cavallaro“CAPE-VLSI Implementation of a Systolic Processor Array: Architecture, Design and Testing,” Proc. 9th Biennial University/Government/Industry Microelectronics Symp., pp.64–69,Jun. 1991.
[30] Weiwei Ma, M. E. Kaye, D. M. Luke and R. Doraiswami, “An FPGA-Based Singular Value Decomposition Processor,” Proc. Canadian Conf. on CCECE, pp. 1047–1050,May. 2006.
[31] P. M. Szecowka and P. Malinowski “CORDIC and SVD Implementationin Digital Hardware,” Proc. Int. Conf. MixedDesign of Integrated Circuits and Systems (MIXDES), pp.237–242, Jun. 2010.
[32] R. P. Brent, F. T.Luk, and C. Van Loan, “Computation of the Singular Value Decomposition Using Mesh-Connected Processors,” J. VLSI Comput. Syst., vol. 1, pp.242–270 1984.
[33] S. G. Mallat and Z. Zhang, “Matching Pursuits with Time-Frequency Dictionaries,” IEEE Trans. Signal Processing, vol. 41, no. 12, pp. 3397–3415, Dec. 1993.
[34] G. Zhang, F. Jiang, D. Zhao, X. Sun and S. Liu “Saliency Detection: A Self-adaption Sparse Representation Approach,” Proc. Int’l Conf. on Image and Graphics (ICIG), pp. 461–465, Aug. 2011.
[35] N. Bruce and J. Tsotsos, “Saliency Based on Information Maximization,” in Advances in Neural Information Processing Systems,(NIPS), pp. 155–162, 2006.
[36] X. Hou and L. Zhang, “Dynamic Visual Attention: Searching for Coding Length Increments,” in Advances in Neural Information Processing Systems, pp. 681–688, 2008.
[37] L. Zhang, M. Tong, T. Marks, H. Shan, and G. Cottrell, “Sun: A Bayesian Framework for Saliency Using Natural Statistics,” Journal of Vision, vol. 8, no. 7, pp. 1–20, Dec. 2008.

指導教授

王家慶(Jia-ching Wang)

審核日期

2014-8-21

推文