本文選題：人臉識(shí)別 + 特征提取�。� 參考：《南昌航空大學(xué)》2017年碩士論文

【摘要】：在人臉識(shí)別過(guò)程中,特征提取的重點(diǎn)在于挖掘并提取人臉數(shù)據(jù)中的關(guān)鍵特征,這有利于提高算法的識(shí)別和分類能力。傳統(tǒng)基于子空間學(xué)習(xí)的特征提取算法如主成分分析(PCA)和線性判別分析(LDA),以及基于流形學(xué)習(xí)的圖嵌入特征提取算法如局部線性嵌入(LLE)和局部保持投影(LPP),因?yàn)榫哂泻?jiǎn)單、直觀、高效等優(yōu)點(diǎn)被廣泛使用。但是上述算法仍然存在許多問(wèn)題和局限性,例如不能同時(shí)得到數(shù)據(jù)的全局和局部結(jié)構(gòu)、線性方法對(duì)于非線性數(shù)據(jù)處理不理想、“小樣本”問(wèn)題以及特征冗余等等�；谙∈杼卣魈崛〉难芯渴侨四樧R(shí)別領(lǐng)域中的另一個(gè)熱點(diǎn)。原始人臉數(shù)據(jù)中往往包含眾多特征,稀疏特征提取可以從原始數(shù)據(jù)中找到某些最顯著的特征,然后使用它們組成最小特征子集對(duì)原始數(shù)據(jù)進(jìn)行最優(yōu)表示,這一過(guò)程既可以簡(jiǎn)化數(shù)據(jù)又能夠保留數(shù)據(jù)中的關(guān)鍵信息。彈性網(wǎng)絡(luò)回歸(Elastic Net)是目前常用的稀疏特征提取算法之一。本文結(jié)合常用圖嵌入算法和彈性網(wǎng)絡(luò)回歸,針對(duì)上述特征提取算法中存在的問(wèn)題進(jìn)行研究,提出新的算法并應(yīng)用在人臉識(shí)別中,主要工作有:(1)簡(jiǎn)單介紹了人臉識(shí)別的研究背景及發(fā)展歷程、研究?jī)?nèi)容及應(yīng)用、存在的問(wèn)題等,并對(duì)幾種典型的人臉數(shù)據(jù)庫(kù)作了簡(jiǎn)要說(shuō)明;(2)根據(jù)本文研究的內(nèi)容,分別介紹了基于流形學(xué)習(xí)的圖嵌入以及稀疏特征提取的思想,并對(duì)經(jīng)典特征提取算法(PCA、LDA、LLE和LPP)以及稀疏特征提取算法(嶺回歸(Ridge)、套索回歸(Lasso)和彈性網(wǎng)絡(luò)回歸)的實(shí)現(xiàn)步驟進(jìn)行了細(xì)致的介紹,然后簡(jiǎn)單分析了上述算法的優(yōu)缺點(diǎn);(3)結(jié)合PCA、LLE以及彈性網(wǎng)絡(luò)回歸,提出了無(wú)監(jiān)督稀疏差分嵌入(USDE)特征提取算法。該算法的基本思想是:首先,構(gòu)建出基于LLE的“局部最小嵌入”以及基于PCA的“全局最大方差”;然后,使用“差分”形式解決多目標(biāo)最優(yōu)化問(wèn)題,并結(jié)合稀疏約束構(gòu)建USDE目標(biāo)函數(shù);最后,使用彈性網(wǎng)絡(luò)回歸進(jìn)行稀疏性實(shí)現(xiàn);(4)在最大邊界準(zhǔn)則(MMC)算法的基礎(chǔ)上,結(jié)合LLE和彈性網(wǎng)絡(luò)回歸提出了基于最大邊界準(zhǔn)則的稀疏局部嵌入(SLE/MMC)算法。首先,SLE/MMC在保持局部近鄰的基礎(chǔ)上構(gòu)建類內(nèi)散布矩陣以及類間散布矩陣;然后,SLE/MMC使用“MMC”的形式以及稀疏約束構(gòu)造SLE/MMC的目標(biāo)函數(shù);最后,SLE/MMC使用彈性網(wǎng)絡(luò)回歸得到一個(gè)稀疏化的結(jié)果。(5)結(jié)合二維判別局部保持投影(2DDLPP)和彈性網(wǎng)絡(luò)回歸,提出了基于稀疏二維判別局部保持投影(S2DDLPP)的特征提取方法。2DDLPP在LPP中引入類間離散度和類別信息,并直接利用原始人臉數(shù)據(jù)矩陣而不是變換后的向量進(jìn)行特征映射,可以減少在變換過(guò)程中的信息損失。首先,在2DDLPP基礎(chǔ)上,S2DDLPP在滿足“類內(nèi)距離最小化”和“類間距離最大化”的同時(shí),在其目標(biāo)函數(shù)上加入稀疏約束;然后,S2DDLPP使用彈性網(wǎng)絡(luò)回歸進(jìn)行稀疏性實(shí)現(xiàn),得到一個(gè)最優(yōu)稀疏投影矩陣。
[Abstract]:In the process of face recognition, the emphasis of feature extraction is to mine and extract the key features from face data, which is helpful to improve the recognition and classification ability of the algorithm. Traditional feature extraction algorithms based on subspace learning, such as principal component analysis (PCA) and linear discriminant analysis (LDAA), and manifold learning based graph embedding feature extraction algorithms, such as local linear embedding (LLEE) and locally preserving projection (LPP), are simple and intuitive. The advantages of high efficiency are widely used. However, these algorithms still have many problems and limitations, such as the global and local structure of the data can not be obtained simultaneously, the linear method is not ideal for dealing with nonlinear data, the "small sample" problem and the feature redundancy and so on. The research of sparse feature extraction is another hotspot in the field of face recognition. Primitive face data often contain many features. Sparse feature extraction can find some of the most prominent features from the original data, and then use them to form a minimal feature subset to represent the original data optimally. This process can both simplify the data and retain the key information in the data. Elastic Networked is one of the commonly used sparse feature extraction algorithms. In this paper, based on the commonly used graph embedding algorithm and elastic network regression, the problems existing in the above feature extraction algorithms are studied, and a new algorithm is proposed and applied to face recognition. The main work is: (1) briefly introduces the research background and development course, research contents and applications, existing problems of face recognition, and gives a brief description of several typical face databases. The ideas of graph embedding and sparse feature extraction based on manifold learning are introduced respectively. The implementation steps of the classical feature extraction algorithms (PCALDALE and LPP) and sparse feature extraction algorithms (Ridgeg Ridge, Lassoand Elastic Network regression) are introduced in detail. Then, the advantages and disadvantages of the above algorithms are briefly analyzed. Combined with PCALLE and elastic network regression, an unsupervised sparse difference embedding (USDE) feature extraction algorithm is proposed. The basic ideas of the algorithm are as follows: firstly, the "local minimum embedding" based on LLE and the "global maximum variance" based on PCA are constructed; then, the "difference" form is used to solve the multi-objective optimization problem. The objective function of USDE is constructed with sparse constraints. Finally, using elastic network regression to implement sparsity, we use the maximum boundary criterion (MMC) algorithm. Combined with LLE and elastic network regression, a sparse local embedding (SLER / MMC) algorithm based on the maximum boundary criterion is proposed. Firstly, the intraclass dispersion matrix and the inter-class dispersion matrix are constructed on the basis of preserving the local nearest neighbor, and then the SLER / MMC uses the form of "MMC" and sparse constraints to construct the objective function of SLE/MMC. Finally, SLER / MMC uses elastic network regression to obtain a sparse result. 5) combining with 2-D discriminant local preserving projection 2DDLPP) and elastic network regression, This paper presents a feature extraction method based on sparse two-dimensional discriminant locally preserving projection (S2DDLPP). 2DDLPP introduces inter-class dispersion and class information into LPP, and directly uses primitive face data matrix instead of transformed vector to map features. It can reduce the loss of information in the process of transformation. Firstly, on the basis of 2DDLPP, S2DDLPP not only satisfies "minimization of intra-class distance" and "maximization of inter-class distance", but also adds sparse constraints to its objective function, then S2DDLPP uses elastic network regression to implement sparsity. An optimal sparse projection matrix is obtained.
【學(xué)位授予單位】：南昌航空大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 劉柳;陶大程;;Lasso問(wèn)題的最新算法研究[J];數(shù)據(jù)采集與處理;2015年01期

2 張哲;梁馮珍;;基于彈性網(wǎng)回歸的居民消費(fèi)價(jià)格指數(shù)分析[J];哈爾濱商業(yè)大學(xué)學(xué)報(bào)(自然科學(xué)版);2013年05期

3 吳巾一;周德龍;;人臉識(shí)別方法綜述[J];計(jì)算機(jī)應(yīng)用研究;2009年09期

4 馬瑞;王家^，

本文編號(hào)：1778890