Online learning from local features for video-based face recognition

This paper presents an online learning approach to video-based face recognition that does not make any assumptions about the pose, expressions or prior localization of facial landmarks. Learning is performed online while the subject is imaged and gives near realtime feedback on the learning status. Face images are automatically clustered based on the similarity of their local features. The learning process continues until the clusters have a required minimum number of faces and the distance of the farthest face from its cluster mean is below a threshold. A voting algorithm is employed to pick the representative features of each cluster. Local features are extracted from arbitrary keypoints on faces as opposed to pre-defined landmarks and the algorithm is inherently robust to large scale pose variations and occlusions. During recognition, video frames of a probe are sequentially matched to the clusters of all individuals in the gallery and its identity is decided on the basis of best temporally cohesive cluster matches. Online experiments (using live video) were performed on a database of 50 enrolled subjects and another 22 unseen impostors. The proposed algorithm achieved a recognition rate of 97.8% and a verification rate of 100% at a false accept rate of 0.0014. For comparison, experiments were also performed using the Honda/UCSD database and 99.5% recognition rate was achieved.     

Constructing and applying higher order textons: Estimating breast cancer risk

Texture analysis based on textons is extended by introducing a method for computing textons of arbitrary order. First-, second- and third-order textons are applied to classify screening mammograms as to indicate a low or high risk of breast cancer. First-order textons are found to provide better estimates of breast cancer risk than other orders on their own but the combination of first- and second-order textons outperforms first-order textons alone and other combinations of two orders. Combining all three orders of textons does not improve classification. This example indicates that including higher-order textons has the potential to improve classification performance.     

Artificially intelligent recognition of Arabic speaker using voice print-based local features

Local features for any pattern recognition system are based on the information extracted locally. In this paper, a local feature extraction technique was developed. This feature was extracted in the time–frequency plain by taking the moving average on the diagonal directions of the time–frequency plane. This feature captured the time–frequency events producing a unique pattern for each speaker that can be viewed as a voice print of the speaker. Hence, we referred to this technique as voice print-based local feature. The proposed feature was compared to other features including mel-frequency cepstral coefficient (MFCC) for speaker recognition using two different databases. One of the databases used in the comparison is a subset of an LDC database that consisted of two short sentences uttered by 182 speakers. The proposed feature attained 98.35% recognition rate compared to 96.7% for MFCC using the LDC subset.     

基于局部特征点方法的动作识别研究

利用局部特征点方法进行视频中的动作识别研究,通过对尺度空间理论以及多种经典的局部特征点检测方法的深入分析,将SIFT算法引入到视频研究领域,提出了一种全新而高效的视频特征点检测算法。在此基础上,设计了合理的机器学习方法对局部特征点进行训练,利用训练出的通用动作模式在视频中进行动作识别。     

基于核函数的高阶特征表示方法

为了解决高阶局部特征带来的计算复杂度提高问题,提出一种基于核函数的高阶局部特征表示方法。通过在两幅图像的局部特征之间进行比较,将特征空间映射到几何不变空间,统计高阶局部特征构建核函数,并结合支持向量机进行多类目标图像分类实验。实验结果分析表明,该方法在提高分类准确率的同时,所需的计算时间只与局部特征的个数呈线性增长。     

基于局部信息统计的人耳识别方法

提出了一种基于局部信息统计的人耳识别方法。该方法将一幅人耳图像分成若干个子区域,分别提取每个子区域的分类特征,将各个子区域的特征串联为一个特征向量构筑人耳特征矢量,更加全面描述了人耳图像的局部与结构信息,应用最近邻分类器进行模式分类。采用三种不同的特征提取方法,以USTB人耳图像库对算法进行测试,实验结果表明,与全局信息比较同种方法识别率提高30%以上,验证了局部信息方法的有效性。     

Handwritten digit classification using higher order singular value decomposition

In this paper we present two algorithms for handwritten digit classification based on the higher order singular value decomposition (HOSVD). The first algorithm uses HOSVD for construction of the class models and achieves classification results with error rate lower than 6%. The second algorithm uses the HOSVD for tensor approximation simultaneously in two modes. Classification results for the second algorithm are almost down at 5% even though the approximation reduces the original training data with more than 98% before the construction of the class models. The actual classification in the test phase for both algorithms is conducted by solving a series least squares problems. Considering computational amount for the test presented the second algorithm is twice as efficient as the first one.     

融合全局和局部特征的Fisherfaces方法

提出了一种融合全局和局部特征的Fisherfaces方法。在Fisher线性准则下,抽取出图像全局特征和局部特征的最佳分类特征。计算待识别样本和训练样本集的加权欧氏距离。在最近邻准则下,判别待识别样本的类别,在ORL人脸库上进行的对比实验结果表明该方法的优越性。     

Geometrical criteria on the higher order smoothness of composite surfaces

A generalization of a theorem by Pegna and Wolter—called Linkage Curve Theorem—is presented. The new theorem provides a condition for joining two surfaces with high order geometric continuity of arbitrary degree n. It will be shown that the Linkage Curve Theorem can be generalized even for the case when the common boundary curve is only G¹.     

On the number of higher order Delaunay triangulations

Higher order Delaunay triangulations are a generalization of the Delaunay triangulation that provides a class of well-shaped triangulations, over which extra criteria can be optimized. A triangulation is order-k Delaunay if the circumcircle of each triangle of the triangulation contains at most k points. In this paper we study lower and upper bounds on the number of higher order Delaunay triangulations, as well as their expected number for randomly distributed points. We show that arbitrarily large point sets can have a single higher order Delaunay triangulation, even for large orders, whereas for first order Delaunay triangulations, the maximum number is 2ⁿ⁻³. Next we show that uniformly distributed points have an expected number of at least 2^{ρ₁n(1+o(1))} first order Delaunay triangulations, where ρ₁ is an analytically defined constant (ρ₁≈0.525785), and for k>1, the expected number of order-k Delaunay triangulations (which are not order-i for any i<k) is at least 2^{ρ_kn(1+o(1))}, where ρ_k can be calculated numerically.     

Learning the best subset of local features for face recognition

We propose a novel, local feature-based face representation method based on two-stage subset selection where the first stage finds the informative regions and the second stage finds the discriminative features in those locations. The key motivation is to learn the most discriminative regions of a human face and the features in there for person identification, instead of assuming a priori any regions of saliency. We use the subset selection-based formulation and compare three variants of feature selection and genetic algorithms for this purpose. Experiments on frontal face images taken from the FERET dataset confirm the advantage of the proposed approach in terms of high accuracy and significantly reduced dimensionality.     

Robust place recognition based on omnidirectional vision and real-time local visual features for mobile robots

Topological localization is especially suitable for human–robot interaction and robot’s high level planning, and it can be realized by visual place recognition. In this paper, bag-of-features, a popular and successful approach in pattern recognition community, is introduced to realize robot topological localization. By combining the real-time local visual features proposed by ourselves for omnidirectional vision and support vector machines, a robust and real-time visual place recognition algorithm based on omnidirectional vision is proposed. The panoramic images from the COLD database were used to perform experiments to determine the best algorithm parameters and the best training condition. The experimental results show that the robot can achieve robust topological localization with high successful rate in real time by using our algorithm.     

基于降维LBP与叶片形状特征的植物叶片识别方法

为解决由于植物叶片特征的相似性以及叶片旋转导致植物识别率较低的问题,提出一种基于降维局部二值模式（LBP）与叶片形状特征相结合的植物叶片识别方法。首先利用LBP算法提取高维叶片纹理信息,通过主成分分析方法（PCA）对高维叶片特征降维;同时考虑叶片的形状特征,将LBP旋转不变性特征与叶片形状特征有效结合,在低维空间利用k近邻法（KNN）实现叶片的分类与识别。实验结果表明该方法具有较好的识别效果。     

Cylindrical bending of piezoelectric laminates with a higher order shear and normal deformation theory

An analytical solution, based on a higher order shear and normal deformation theory, is presented for the cylindrical flexure of piezoelectric plates. The primary displacement terms are expanded in thickness coordinate and an exact nature of electric potential is obtained in actuator and sensing layers. The electric potential function is evaluated by solving a second order ordinary differential equation satisfying electric boundary conditions along thickness direction of piezoelectric layer. A unidirectional composite plate attached with distributed actuator and sensor layers is analyzed under electrical and mechanical loading conditions and comparison of results with exact solution is presented. Results for non-piezoelectric plates are also compared with elasticity and other solutions of cylindrical bending.     

Numerical methods of higher order of accuracy for incompressible flows

The work deals with numerical solution of the Navier–Stokes equations for incompressible fluid using finite volume and finite difference methods. The first method is based on artificial compressibility where continuity equation is changed by adding pressure time derivative. The second method is based on solving momentum equations and the Poisson equation for pressure instead of continuity equation. The numerical solution using both methods is compared for backward facing step flows. The equations are discretized on orthogonal grids with second, fourth and sixth orders of accuracy as well as third order accurate upwind approximation for convective terms. Not only laminar but also turbulent regimes using two-equation turbulence models are presented.     

Fusion of local normalization and Gabor entropy weighted features for face identification

Face recognition is one of the most extensively studied topics in image analysis because of its wide range of possible applications such as in surveillance, access control, content-based video search, human–computer interaction, electronic advertisement and more. Face identification is a one-to-n matching problem where a captured face is compared to n samples in a database. In this work we propose two new methods for face identification. The first one combines entropy-like weighted Gabor features with the local normalization of Gabor features. The second fuses the entropy-like weighted Gabor features at the score level with the local binary pattern (LBP) applied to the magnitude (LGBP) and phase (LGXP) components of the Gabor features. We used the FERET, AR, and FRGC 2.0 databases to test and compare our results with those previously published. Results on these databases show significant improvement relative to previously published results, reaching the best performance on the FERET and AR databases. Our methods also showed significant robustness to slight pose variations. We tested the proposed methods assuming noisy eye detection to check their robustness to inexact face alignment. Results show that the proposed methods are robust to errors of up to 3 pixels in eye detection.     

Optimal higher order learning adaptive control approach for a class of SISO nonlinear systems

1IntroductionMany dynamic systems to be controlled have constant orslowly_varying uncertain parameters .Adaptive control is apopular approachtothe control ofsuchsystems [1] .Inthepast two decades ,significant progress has been made in theresearch and design of adaptive control systems [2,3] .Fairly complete and comprehensive guidelines are nowavailable for both design and implementation of adaptivecontrollers inthe cases where the systemunder control canbe adequately modeled as a linear dynami…     

Optimal higher order learning adaptive control approach for a class of SISO nonlinear systems

In this paper,an optimal higher order learning adaptive control approach is developed for a class of SISO nonlinear systems.This design is model-free and depends directly on pseudo-partial-derivatives derived on-line from the input and output information of the system.A novel weighted one-step-ahead control criterion function is proposed for the control law.The convergence analysis shows that the proposed control law can guarantee the convergence under the assumption that the desired output is a set point.Simulation examples are provided for nonlinear systems to illustrate the better performance of the higher order learning adaptive control.     

局部二值描述子的研究进展综述

局部二值描述子是局部不变特征中的重要研究对象,广泛应用于计算机视觉与模 式识别中。近年来,以 BRIEF 描述子为代表的局部二值描述子相继出现,对十年来局部二值描 述子的研究成果与发展方向进行综述,旨在为初步研究者与工程应用人员提供参考。首先,对 典型的现代局部二值描述子进行概述;其次,对优化局部二值描述子方法进行分析;最后,对 相关实验评估准则进行讨论,通过总结现阶段存在的问题,给出未来研究的展望。从整体来看, 近年来局部二值描述子经历了显著的发展与进步,许多对于局部二值描述子的研究均在普适性、 鲁棒性和高效性上取得了成果。针对应用场景的不同,部分优化后的描述子也具备了应对实际 问题的能力。这些研究进展为局部二值描述子向高层次发展、多领域拓宽打下了坚实的基础并 提供了更多的思路。局部二值描述子的成功发展标志着计算机视觉技术的进步,但其发展过程 中依然存在一些共性问题与矛盾,有待进一步的深入研究与解决。     

融合语音情感词局部特征的语音情感识别方法

为有效利用语音情感词局部特征,提出了一种融合情感词局部特征与语音语句全局特征的语音情感识别方法。该方法依赖于语音情感词典的声学特征库,提取出语音语句中是否包含情感词及情感词密度等局部特征,并与全局声学特征进行融合,再通过机器学习算法建模和识别语音情感。对比实验结果表明,融合语音情感词局部特征与全局特征的语音情感识别方法能取得更好的效果,局部特征的引入能有效提高语音情感识别准确率。