av无码久久久久不卡免费网站,99久久婷婷国产综合精品草原,久久久精品国产

matlab函數fminunc:求無約束優化的最優解

例：K-means Hashing提供的代碼optimize_one_iter函數內部用到fminunc，代價函數是eval_one_center函數，看eval_one_center函數的注釋是this function computes the cost of Eqn. (8) given a center c_j

posted @ 2015-08-13 12:59 杰哥閱讀(877) | 評論 (0) | 編輯收藏

[zz] 機器視覺開源代碼集合（轉載）

摘要: http://xilinx.eetrend.com/article/8919 一、特征提取Feature Extraction： SIFT [1] [Demo program][SIFT Library] [VLFeat]PCA-SIFT [2] [Project]Affine-SIFT [3] [Project]SURF [4] [OpenSURF] [Matlab Wrapper]Affi... 閱讀全文

posted @ 2015-08-13 10:19 杰哥閱讀(1196) | 評論 (0) | 編輯收藏

計算機視覺-機器學習領域近兩年部分綜述文章

http://www.valseonline.org/thread-505-1-1.html

【2015】

[1]. E.Sariyanidi, H. Gunes, A. Cavallaro, Automatic Analysisof Facial Affect: A Survey of Registration, Representation, and Recognition,IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 37, NO. 6,JUNE 2015

2015-1. Automatic Analysis of Facial Affect A Survey of Registration, Representa.pdf(1.51 MB, 下載次數: 6)

[2]. T. Li,H. Chang, M. Wang, B.B. Ni, R.C. Hong, S.C. Yan, CrowdedScene Analysis: A Survey, IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FORVIDEO TECHNOLOGY, VOL. 25, NO. 3, MARCH 2015

2015-2. Crowded Scene Analysis A Survey.pdf(5.42 MB, 下載次數: 19)

[3]. Z.Zhang, Y. Xu, J. Yang, X.L. Li, D. Zhang, A Survey ofSparse Representation: Algorithms and Applications, IEEE ACCESS, date ofpublication May 6, 2015

2015-3. A Survey of Sparse Representation.pdf(4.8 MB, 下載次數: 0)

[4]. J.Galbally, S. Marcel, J. Fierrez, Biometric AntispoofingMethods: A Survey in Face Recognition, IEEE ACCESS, date of publicationDecember 18, 2014

2015-4. Biometric Antispoofing Methods A Survey in Face Recognition.pdf(10.43 MB, 下載次數: 0)

[5]. B.Tian, B. T. Morris, M. Tang, Y.Q. Liu, Y. J. Yao, C. Guo, D.Y. Shen, S.H. Tang, Hierarchical and Networked Vehicle Surveillance in ITS:A Survey, IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, VOL.16, NO. 2, APRIL 2015

2015-5. Hierarchical and Networked Vehicle Surveillance in ITS A Survey.pdf(1.12 MB, 下載次數: 0)

[6]. A. Betancourt,P. Morerio, C. S. Regazzoni, and M. Rauterberg, TheEvolution of First Person Vision Methods: A Survey, IEEE TRANSACTIONS ONCIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 25, NO. 5, MAY 2015

2015-6. The Evolution of First Person Vision Methods A Survey.pdf(2.69 MB, 下載次數: 0)

[7]. L.Shao, F. Zhu, and X.L. Li, Transfer Learning for VisualCategorization: A Survey, IEEE TRANSACTIONS ON NEURAL NETWORKS ANDLEARNING SYSTEMS, VOL. 26, NO. 5, MAY 2015

2015-7. Transfer Learning for Visual Categorization A Survey.pdf(3.11 MB, 下載次數: 0)

【2014】

[8]. S. Fu,H. B. He, Z.G. Hou, Learning Race from Face: A Survey, IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 36, NO.12, DECEMBER 2014

2014-1. Learning Race from Face A Survey.pdf(2.55 MB, 下載次數: 15)

[9]. H.L. Zhou,A. Mian, L. Wei, D. Creighton, M. Hossny, and S. Nahavandi, Recent Advances on Singlemodal and Multimodal FaceRecognition: A Survey, IEEE TRANSACTIONS ON HUMAN-MACHINE SYSTEMS, VOL.44, NO. 6, DECEMBER 2014

2014-2. Recent Advances on Singlemodal and Multimodal Face Recognition A Survey.pdf(669.19 KB, 下載次數: 0)

【2013】

[10]. O. D. Lara, M.A. Labrador, A Survey on Human Activity Recognition using WearableSensors, IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 15, NO. 3,THIRD QUARTER 2013

2013-1. A Survey on Human Activity Recognition using Wearable Sensors.pdf(1.65 MB, 下載次數: 8)

[11]. A. Sotiras, C. Davatzikos, Nikos. Paragios, Deformable Medical Image Registration: A Survey, IEEETRANSACTIONS ON MEDICAL IMAGING, VOL. 32, NO. 7, JULY 2013

2013-2.Deformable Medical Image Registration A Survey.pdf(1.3 MB, 下載次數: 11)

[12]. A. Alrahayfeh, M. Faezipour, Eye Tracking and Head Movement Detection: A State-of-ArtSurvey, IEEE Journal of Translational Engineering in Health andMedicine, 2013

2013-3. Eye Tracking and Head Movement Detection A State-of-Art Survey.pdf(2.22 MB, 下載次數: 5)

[13]. P.V.K. Borges, N. Conci, and A. Cavallaro, Video-Based Human Behavior Understanding: A Survey, IEEETRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 23, NO. 11,NOVEMBER 2013

2013-4. Video-Based Human Behavior Understanding A Survey.pdf(15.27 MB, 下載次數: 46)

posted @ 2015-08-12 12:09 杰哥閱讀(255) | 評論 (0) | 編輯收藏

如何分析這兩個集合的分布的相似程度？

降維(比如PCA或者random projection) -> KDE(kernel density estimation)來估算密度 -> KL divergence

posted @ 2015-08-09 13:05 杰哥閱讀(415) | 評論 (0) | 編輯收藏

matlab加速: bsxfun

%將訓練數據和測試數據都去中心化
X = traindata;
label = traingnd;
m = mean(X);
X_zm = bsxfun(@minus, X, m);
traindata_zm = bsxfun(@minus, traindata, m);
testdata_zm = bsxfun(@minus, testdata, m);

matlab函數 bsxfun淺談（轉載）
http://blog.sina.com.cn/s/blog_9e67285801010ttn.html

網上關于bsxfun的東西不多，今天需要看到一個，由于原博文插入的圖片顯示不出來，于是筆者大發善心進行了contrl+V 以及alt+ctrl+A的操作，供大家交流學習。

bsxfun是一個matlab自版本R2007a來就提供的一個函數，作用是”applies an element-by-element binary operation to arrays a and b, with singleton expansion enabled.”

舉個例子。假設我們有一列向量和一行向量。

a = randn(3,1), b = randn(1,3) a = -0.2453 -0.2766 -0.1913 b = 0.6062 0.5655 0.9057

我們可以很簡單的使用matlab的外乘c=a*b來得到，如圖 bsxfun淺談（轉載）" o:button="t" target='"_blank"' o:spid="_x0000_i1025">bsxfun淺談（轉載）" src="file:///C:\Users\jie\AppData\Local\Temp\msohtmlclip1\01\clip_image001.jpg">
但如果我們想用”外加”呢？也就是說把上式求解過程中的乘號換做加號？

這時我們可以用c=bsxfun(@plus,a,b)來實現。

bsxfun的執行是這樣的，如果a和b的大小相同，那么c=a+b. 但如果有某維不同，且a或b必須有一個在這一維的維數為1, 那么bsxfun就將少的這個虛擬的復制一些來使與多的維數一樣。在我們這里，b的第一維只有1（只一行），所以bsxfun將b復制3次形成一個3×3的矩陣，同樣也將a復制成3×3的矩陣。這個等價于c=repmat(a,1,3)+repmat(b,3,1)。這里

repmat(a,1,3) ans = -0.2453 -0.2453 -0.2453 -0.2766 -0.2766 -0.2766 -0.1913 -0.1913 -0.1913

repmat是顯式的復制，當然帶來內存的消耗。而bsxfun是虛擬的復制，實際上通過for來實現，等效于for(i=1:3),for(j=1:3),c(i,j)=a(i)+b(j);end,end。但bsxfun不會有使用matlab的for所帶來額外時間。實際驗證下這三種方式

>> c = bsxfun(@plus,a,b) c = 0.3609 0.3202 0.6604 0.3296 0.2889 0.6291 0.4149 0.3742 0.7144 >> c = repmat(a,1,3)+repmat(b,3,1) c = 0.3609 0.3202 0.6604 0.3296 0.2889 0.6291 0.4149 0.3742 0.7144 >> for(i=1:3),for(j=1:3),c(i,j)=a(i)+b(j);end,end,c c = 0.3609 0.3202 0.6604 0.3296 0.2889 0.6291 0.4149 0.3742 0.7144

從計算時間上來說前兩種實現差不多，遠高于for的實現。但如果數據很大，第二種實現可能會有內存上的問題。所以bsxfun最好。

下面看一個更為實際的情況。假設我們有數據A和B, 每行是一個樣本，每列是一個特征。我們要計算高斯核，既：

這里@plus是加法的函數數柄，相應的有減法@minus, 乘法@times, 左右除等，具體可見 doc bsxfun.

k(||x-xc||)=exp{- ||x-xc||^2/(2*σ)^2) } 其中xc為核函數中心,σ為函數的寬度參數 , 控制了函數的徑向作用范圍。

當然可以用雙重for實現（如果第一直覺是用三重for的話…）。