動作識別在各種領域上有越來越活躍的發展，運用範圍極為廣闊，從國土保全、財物人身的保障，到居家照護、甚至是智慧環境、體感遊戲等等都是其範疇。本篇論文針對嵌入式系統上進行動作識別的演算法進行分析，發現許多區塊重複地進行相同的運算，此類型的運算可以被併行處理，並利用SIMD 的架構來加速。本論文嘗試用多串流多媒體處理單元(MSMU)來實現動作識別演算法，MSMU 架構是一個類MMX 的SIMD 處理架構，本身包含了SIMD 運算以及與暫存空間兩種功能，並引入多資料流的處理概念，可藉由模式的切換來動態的調整資料流的並行度。藉由的模式切換，與新增的轉置指令來處理平面的運算，並探討模式切換所帶來的好處。藉由比較128-bit 的SSE架構與MSMU 在處理一些實際的例子上，凸顯單純增加subword 的並行度所面臨的問題，顯現出多資料流帶來的優勢。針對演算法的部分，研究以切割SIMD 最小的元素以及使用全位元運算子的方式來提高運算的並行度，以達到更好的效率提升。MSMU 與現有的嵌入式SIMD 架構WMMX 相比，可以達到3.49 倍的提升。

Action recognition had become prosperous in development and been broadly applied in several sectors. From homeland security, personal property, home caring, even the smart environment and the motion-sensing games, are in its territories
This paper analysis the algorithm of Action recognition for embedded system, finds that there are many blocks can use the parallel execution to compute more efficiently. This paper tries to implement action recognition algorithm on Multiple-Streaming Multimedia Unit (MSMU). MSMU is a MMX-like SIMD architecture, with SIMD Operation and Data Storage. By introduction the concept of multiple streaming, MSMU will be able to modulate the amount of parallel data streams dynamically via switching the instruction mode. With Mode Switching and new added transfer instruction to compute 2D image processing, study the benefit of the instruction mode switching
Through comparing the 128-bit SSE architecture and MSMU architecture with the practical example, highlight the problems that exploiting the subword parallelisms facing and bring out the advantage of Multistreaming.
For the algorithm, study the slicing the minimum element and using the bitwise operation approach to better efficiency. Compare to embedded SIMD architecture "WMMX", MSMU can achieve 3.49× overall speedup.

摘要 I
ABSTRACT II
圖目錄 VII
表目錄 X
第一章 簡介 1
1.1 研究動機 1
1.2 研究目的 2
1.3 論文架構 2
第二章 相關研究 3
2.1 動作識別演算研究 3
2.1.1 MGD特徵動作識別演算法研究 4
2.2 Support Vector Machine 支持向量機 9
2.3 MSMU (Multiple-Streaming Multimedia Unit) 多串流多媒體處理單元 16
2.4 相關SIMD指令集研究(MMX、SSE與WMMX) 23
2.4.1 MMX 23
2.4.2 SSE 23
2.4.3 WMMX 25
第三章 動作識別演算法實現 26
3.1 MGD特徵萃取演算法實現 26
3.1.1 DMASKS演算法實現 26
3.1.2 HMHHb演算法實現 33
3.1.3 MGD演算法實現 40
3.2 支持向量機實現 51
第四章 模擬平台的建立與實現 54
4.1 MSMU模擬平台的建構 54
4.2 WMMX模擬平台的建構 55
4.3 測試影片資料庫 56
4.4 支持向量機實現 57
4.4.1 支持向量模型建立 58
4.4.2 支持向量機的分類 59
第五章 實驗結果與分析 60
5.1 MGD動作識別的實現 60
5.2 MGD以SIMD架構實現 62
5.2.1 DMASKS萃取 62
5.2.2 HMHH萃取 63
5.2.3 MGD萃取 65
5.2.4 SVM分類函數 65
5.3 總體效能分析 67
第六章 結論 69
參考資料 70

[1] Hongying Meng, Nick Pears, Chris Bailey, “A Human Action Recognition System for Embedded Computer Vision Application,” IEEE Conference on Computer Vision and Pattern Recognition Minneapolis, MN, USA , June 2007.
[2]Hongying Meng , Michael Freeman, Nick Pears, Chris Bailey, “Real-time human action recognition on an embedded, reconfigurable video processing architecture,” Journal of Real-Time Image Processing , pp. 163-176, Sep. 2008 .
[3]Ivan Laptev, Patrick P′erez, “Retrieving actions in movies,” IEEE International Conference on Computer Vision , pp. 1-8 ,Oct 2007.
[4] Lena Gorelick, Moshe Blank,Eli Shechtman, Michal Irani, Ronen Basri, “Actions as Space-Time Shapes,” IEEE Transactions On Pattern Analysis And Machine Intelligence VOL. 29, NO. 12, pp. 2247-2253, Dec. 2007.
[5] Lena Gorelick, Meirav Galun, Eitan Sharon, Ronen Basri, Achi Brandt, “Shape Representation and Classification Using the Poisson Equation,” IEEE Conference on Computer Vision and Pattern Recognition, pp. 1991-2005, Dec 2006.
[6] Yair Shapira ,“Solving PDEs in C++ Numerical Methods in a Unified Object-Oriented Approach,” 2005.
[7] J. W. Davis, “Hierarchical motion history images for recognizing human motion,” In IEEE Workshop on Detection and Recognition of Events in Video, pp. 39–46, 2001.
[8] Meng, N. Pears, and C. Bailey, “Recognizing human actions based on motion information and SVM,” International Conference on Intelligent Environments, pp 239–245, 2006.
[9]Aaron F. Bobick, James W. Davis, “The recognition of human movement using temporal templates,” IEEE Trans. Pattern Anal. Mach. Intell. ,pp. 257-261, Mach 2001.
[10] Chih-Chung Chang , Chih-Jen Lin, “A Practical Guide to Support Vector Classification, ” May 2009.
[11]C. Schuldt, I. Laptev, and B. Caputo, “Recognizing Human Actions: A Local SVM Approach,” Computational Vision and Active Perception Laboratory , Cambridge, UK. 2004.
[12] Jih-Ching Chiu, Yu-Liang Chou, Hua-Yi Tzeng, “A Multi-streaming SIMD Architecture for Multimedia Applications,” Conference On Computing Frontiers Ischia, pp. 51-60, 2009.
[13] Hua-Yi Tzeng, “Implementation of face detection algorithm with parallel extended-MMX instruction set,” July 2008.
[14]Stefano Tommesani, Intel MMX Instruction Set, DOI=http://www.tommesani.com/MMXPrimer.html.
[15]Intel Corp, “MMXT Technology Manuals and Application Notes,” June 2009. DOI=http://software.intel.com/en-us/articles/mmxt-technology-manuals-and-application-notes/
[16]Stefano Tommesani, Intel SSE2 Instruction Set, DOI=http://www.tommesani.com/SSE2Intro.html.
[17]Intel Corp, “Intel® Wireless MMXTM Technology Developer Guide,” Aug. 2002.
[18]Mark Woh, Sangwon Seo, Scott Mahlke, Trevor Mudge, Chaitali Chakrabarti and Krisztian Flautner, “AnySP: Anytime Anywhere Anyway Signal Processing,” International Symposium on Computer Architecture, Austin, TX, USA , pp 128-139 , June 2009 .
[19] Sandeep.S , “GCC-Inline-Assembly-HOWTO,” Mar. 2003. DOI=http://www.ibiblio.org/gferg/ldp/GCC-Inline-Assembly-HOWTO.html
[20]Bryan Catanzaro, Narayanan Sundaram, Kurt Keutzer ,“Fast Support Vector Machine Training and Classication on Graphics Processors,” International Conference on Machine Learning ,pp. 104-111, July 2008.
[21]Chih-Chung Chang , Chih-Jen Lin ,“LIBSVM: a Library for Support Vector Machines,” Feb. 2009.