Image Region Selection and Ensemble for Face Recognition

In this paper, a novel framework for face recognition, namely Selective Ensemble of Image Regions （SEIR）, is proposed. In this framework, all possible regions in the face image are regarded as a certain kind of features. There are two main steps in SEIR： the first step is to automatically select several regions from all possible candidates; the second step is to construct classifier ensemble from the selected regions. An implementation of SEIR based on multiple eigenspaces, namely SEME, is also proposed in this paper. SEME is analyzed and compared with eigenface, PCA ＋ LDA, eigenfeature, and eigenface ＋ eigenfeature through experiments. The experimental results show that SEME achieves the best performance.     

Bilinear Models for 3-D Face and Facial Expression Recognition

In this paper, we explore bilinear models for jointly addressing 3-D face and facial expression recognition. An elastically deformable model algorithm that establishes correspondence among a set of faces is proposed first and then bilinear models that decouple the identity and facial expression factors are constructed. Fitting these models to unknown faces enables us to perform face recognition invariant to facial expressions and facial expression recognition with unknown identity. A quantitative evaluation of the proposed technique is conducted on the publicly available BU-3DFE face database in comparison with our previous work on face recognition and other state-of-the-art algorithms for facial expression recognition. Experimental results demonstrate an overall 90.5% facial expression recognition rate and an 86% rank-1 face recognition rate.      

基于球面调和函数的3-D人脸识别方法

根据3-D数据的优势,利用图像成像原理和球面调和函数理论,结合3-D投影原理和PCA技术建立了一个3-D人脸模型.该模型用不受光线影响的低维线性子空间的基向量来表示,将结构和纹理两个3-D信息作为整体进行考虑,使得模型只需要通过一组参数简单描述.由于本文构建的3-D模型只与人脸的内在属性有关,与光线无关,因此能够排除光线对人脸识别率的影响,本文在AR人脸数据库上的识别实验证明了本文方法的有效性.     

基于集成学习的规范化LDA人脸识别

针对人脸识别问题中经常面临的“小样本”问题,在规范化的LDA算法的基础上加以改进,并结合集成学习的方法,利用Adaboost算法,在每一次的迭代过程中引进一个加权函数对难以分离的样本增加权重。增加分类器之间的差异度,从而提高样本在新的特征空间中的可分离性,将识别率提高至98.5%。通过ORL数据库的大量实验表明,该算法比传统算法有更好的性能。     

A Constraint-Satisfaction Approach for 3-D Object Recognition by Integrating 2-D and 3-D Data

We present a framework for recognition of 3D objects by integrating 2D and 3D sensory data. The major thrust of this work is to efficiently utilize all relevant data, both 2D and 3D, in the early stages of recognition, in order to reduce the computational requirements of the recognition process. To achieve this goal, we formulate the problem as a constraint–satisfaction problem (CSP). Rather than directly solving the CSP, a problem of exponential complexity, we only enforce local consistency in low-order polynomial time. This step of local-consistency enforcement can significantly decrease the computational load on subsequent recognition modules by (1) significantly reducing the uncertainty in the correspondence between scene and model features and (2) eliminating many erroneous model objects and irrelevant scene features. A novel method is presented for efficiently constructing a CSP corresponding to a combination of 2D and 3D scene features. Performance of the proposed framework is demonstrated using simulated and real experiments involving visual (2D) and tactile (3D) data.     

基于模糊积分的不完全小波包子空间集成人脸识别

提出一种基于模糊积分的不完全小波包子空间集成人脸识别方法,并与五种相关方法进行实验比较.首先对人脸图像做不完全小波包分解,对双向低频子空间图像直接进行特征提取,对含有一个方向低频成分的高频子空间图像先求平均,再进行提取特征;然后用得到的不同子空间图像训练模糊分类器;最后用模糊积分融合训练的模糊分类器.该方法能够充分利用不同频率小波子空间图像中包含的有用信息,从而提高人脸识别的精度.在ORL、YALE、JAFFE和FERET这4个人脸数据库上进行实验,实验结果表明该方法在识别精度方面均优于五种相关方法.     

三维人脸识别研究综述

近二十多年来,虽然基于图像的人脸识别已取得很大进展,并可在约束环境下获得很好的识别性能,但仍受光照、姿态、表情等变化的影响很大,其本质原因在于图像是三维物体在二维空间的简约投影.因此,利用脸部曲面的显式三维表达进行人脸识别正成为近几年学术界的研究热点.文中分析了三维人脸识别的产生动机、概念与基本过程;根据特征形式,将三维人脸识别算法分为基于空域直接匹配、基于局部特征匹配、基于整体特征匹配三大类进行综述;对二维和三维的双模态融合方法进行分类阐述;列出了部分代表性的三维人脸数据库;对部分方法进行实验比较,并分析了方法有效性的原因;总结了目前三维人脸识别技术的优势与困难,并探讨了未来的研究趋势.     

基于粗糙集属性约减和神经网络集成的人脸识别技术*

提出了一种基于粗糙集的属性约简方法对经由PCA处理后的人脸特征进行提取,随后使用一种神经网络集成的方法对约简后的人脸数据进行识别。实验结果表明,使用该方法约简后的数据与使用仅由PCA处理后的数据相比,两者获得的识别精度基本相同,而前者的数据属性维数远小于后者,从而大大降低了识别过程的复杂程度。     

基于三维建模的眼镜遮挡下人脸识别

眼镜作为人脸特征的不稳定性是眼镜遮挡人脸识别的主要问题。为避免现有方法消除不稳定眼镜特征时带来的人脸特征丢失,将眼镜视为人脸固有部分,提出一种基于三维建模生成人脸虚拟样本补偿眼镜不稳定性的方法。三维建模方便眼镜模型参数的调节。通过调节眼镜参数,具体分析眼镜不同部分对人脸识别的影响,同时,针对影响严重的镜片模糊和反光,均做相应处理。CAL-PEAL的实验表明本文方法对识别性能的大幅度改善,并验证镜片处理的有效性。     

Anthropometric 3D Face Recognition

We present a novel anthropometric three dimensional (Anthroface 3D) face recognition algorithm, which is based on a systematically selected set of discriminatory structural characteristics of the human face derived from the existing scientific literature on facial anthropometry. We propose a novel technique for automatically detecting 10 anthropometric facial fiducial points that are associated with these discriminatory anthropometric features. We isolate and employ unique textural and/or structural characteristics of these fiducial points, along with the established anthropometric facial proportions of the human face for detecting them. Lastly, we develop a completely automatic face recognition algorithm that employs facial 3D Euclidean and geodesic distances between these 10 automatically located anthropometric facial fiducial points and a linear discriminant classifier. On a database of 1149 facial images of 118 subjects, we show that the standard deviation of the Euclidean distance of each automatically detected fiducial point from its manually identified position is less than 2.54 mm. We further show that the proposed Anthroface 3D recognition algorithm performs well (equal error rate of 1.98% and a rank 1 recognition rate of 96.8%), out performs three of the existing benchmark 3D face recognition algorithms, and is robust to the observed fiducial point localization errors.     

基于改进卷积神经网络与集成学习的人脸识别算法

针对复杂卷积神经网络(CNN)在中小型人脸数据库中的识别结果容易出现过拟合现象,提出一种基于改进CNN网络与集成学习的人脸识别算法。改进CNN网络结合平面网络和残差网络的特点,采用平均池化层代替全连接层,使得网络结构简单且可移植性强。在改进CNN网络的基础上,利用基于投票法的集成学习策略将所有个体学习器结果凸组合为最终结果,实现更准确的人脸识别。实验结果表明,该算法在Color FERET、AR和ORL人脸数据库上的识别准确率分别达到98.89%、99.67%和100%,并且具有较快的收敛速度。     

A Hybrid Features Extraction on Face for Efficient Face Recognition

Multimedia Tools and Applications - Image Processing is one of the vibrant research areas nowadays and particularly face recognition is given much importance in all the sectors. Accordingly this...     

基于集成人脸对距离学习的跨年龄人脸验证*

针对不同年龄跨度下人脸对差异的不同,文中提出基于集成人脸对距离学习(EFPML)的跨年龄人脸验证方法.对不同年龄跨度的人脸对分别学习距离度量,然后使用集成方法对人脸对进行重表示,使人脸对重表示更具有判别性,并且可以扩充有限的跨年龄数据集.在公开的跨年龄人脸数据库FG-NET和CACD上的实验表明,文中方法可以有效减少年龄带来的影响,提高验证性能.     

基于排除算法的快速三维人脸识别方法

提出了一种基于排除算法的快速三维人脸识别方法. 首先, 利用主成分分析(Principal component analysis, PCA)对自动切割的不同姿态人脸进行校正, 将所有人脸转换到统一的坐标系下; 然后提取人脸侧面轮廓线, 利用基于LTS-Hausdorff距离的轮廓线对齐方法对库集对象进行排除; 最后, 采用基于刚性区域的改进迭代最近点(Iterative closest point, ICP)算法对剩余的库集模型进行精确匹配, 给出最终识别结果. 在FRGC V2.0人脸数据库的实验结果表明, 该方法具有较好的实时性和鲁棒性.     

静态人脸识别的一种协同学方法

本文针对静态人脸识别设计了一种简单易行的协同算法,它不需要面部特征的准确配准,当系统中的模式有增减时,算法可以快速获得新的原型向量和伴随向量。此外,对注意参数的两种情况的设置使算法的识别效果较单一平衡注意参数更好。仿真实验证明,该算法具有较好的识别效果。     

基于模型的三维物体识别

实现了一个完整的基于模型的三维物体识别系统,它可识别灰度图象中包含的物体,如对遮挡加以限制,还可识别被遮挡的物体.该系统能实现物体的自动建模,也可先定性识别某一物体的立体图对以获取高层知识,然后在高层知识的指导下准确地匹配立体图对中相对应的特征.此外,还提出了利用最能表示物体特征的表面(特征面)来识别物体的方法,以提高系统抗噪声的能力.大量实验证明,该系统具有相当的稳健性.     

Probabilistic Models of Appearance for 3-D Object Recognition

We describe how to model the appearance of a 3-D object using multiple views, learn such a model from training images, and use the model for object recognition. The model uses probability distributions to describe the range of possible variation in the object's appearance. These distributions are organized on two levels. Large variations are handled by partitioning training images into clusters corresponding to distinctly different views of the object. Within each cluster, smaller variations are represented by distributions characterizing uncertainty in the presence, position, and measurements of various discrete features of appearance. Many types of features are used, ranging in abstraction from edge segments to perceptual groupings and regions. A matching procedure uses the feature uncertainty information to guide the search for a match between model and image. Hypothesized feature pairings are used to estimate a viewpoint transformation taking account of feature uncertainty. These methods have been implemented in an object recognition system, OLIVER. Experiments show that OLIVER is capable of learning to recognize complex objects in cluttered images, while acquiring models that represent those objects using relatively few views.     

3-D Tag-Based RFID System for Recognition of Object

Generally, object recognition in robotics has used vision, sonar, laser range sensors, etc. In this paper, the authors propose a new sensing technology based on radio frequency identification (RFID) to facilitate the autonomous recognition of objects. This technology uses a novel tag, named the 3-D tag, with which an object can be identified and its orientation together with location can be estimated. This paper presents the method for object recognition by analyzing the characteristics of the 3-D tag, and then validates the 3-D tag through several demonstrations showing its usefulness as an independent sensor for recognizing objects.     

人脸识别技术综述

人脸识别技术的研究在近几年得到了高度重视,已经成为图像分析和理解中最成功的应用之一。文中对人脸识别技术的常用方法如：肤色模型方法、基于特征脸(Eiganface)的PCA方法、人工神经网络(ANN)的方法、以及基于局部Haar特征的方法等,进行了分类总结。在分析了影响人脸识别效果的两个主要因素：光照影响和姿态影响后,指出利用先验知识,综合多种分类方法是人脸识别研究的趋势。     

基于形状和纹理的人脸自动识别

人脸识别技术在商业和法律上有广泛的应用前景，在安全监控中也大有用武之地。其主要任务是利用已有的人脸图像库，识别静止的或视频图像中的一张或多张人脸．实验首先提取人脸图像的形状和纹理特征，运用广义KL变换降低形状和纹理空间的维数，避开入脸识别小样本集的局限,同时，通过运用具有统计不相关性的最佳鉴别变换，来抽取人脸图像的有效鉴别特征．在包含960幅人脸图像的NUST603人脸图像库上进行识别实验，得到的识别错误率低于4％，且这种方法对人脸的姿态、表情等条件有一定的不敏感性．