A Novel Face Recognition Method Using Vector Projection

Vector projection for face recognition

This paper presents a new method of human face recognition based on a novel algebraic feature extraction method. An input human face image is First transformed into a standard image; Then, the projective feature vectors of the standard image are extracted by projecting it onto the optimal discriminant projection vectors; Finally, face image recognition is completed by classifying these projective feature vectors. Experimental results showed that the present method is effective.

A method combining eigenface with different wavelet subbands for face recognition is proposed. Each training image is decomposed into multi-subbands for extracting their eigenvector sets and projection vectors. In the recognition process, the inner product distance between the projection vectors of the test image and that of the training image are calculated. The training image, corresponding to the maximum distance under the given threshold condition, is considered as the final result. The experimental results on the ORL and YALE face database show that, compared with the eigenface method directly on the image domain or on a single wavelet subband, the recognition accuracy using the proposed method is improved by 5% without influencing the recognition speed.

Face recognition using locality sensitive histograms of oriented gradients

Person identification in Ethnic Indian Goans using ear biometrics and neural networks

How to extract the robust discrimination features is the key of face recognition (FR). Local binary pattern is one of the most widely used feature extracting method in FR for its comprehensive representation of the visual content of face image. However, the feature vector extracted by LBP is usually very high-dimensional and maybe contains information redundancy. To deal with the drawback of LBP, a novel nonlinear version of LBP is presented. The main idea is firstly all the feature vectors extracted by LBP are mapped into a feature space by a nonlinear mapping, and then the mapped features are expressed using the corresponding projection vectors. Lastly, FR is performed based on the projection vectors. Compared with LBP, the new method has two advantages. Firstly, it can capture the nonlinear information of the feature vector extracted by LBP. Secondly, it avoids the complex expression of the nonlinear mapping. The experimental results on two public standard visual face datasets demonstrate the proposed method is superior to LBP in recognition accuracy while its computational complexity is considerably reduced.

Singular values (SVs) feature vectors of face image have been used for face recognition as the feature recently. Although SVs have some important properties of algebraic and geometric invariance and insensitiveness to noise, they are the representation of face image in its own eigen-space spanned by the two orthogonal matrices of singular value decomposition (SVD) and clearly contain little useful information for face recognition. This study concentrates on extracting more informational feature from a frontal and upright view image based on SVD and proposing an improving method for face recognition. After standardized by intensity normalization, all training and testing face images are projected onto a uniform eigen-space that is obtained from SVD of standard face image. To achieve more computational efficiency, the dimension of the uniform eigen-space is reduced by discarding the eigenvectors that the corresponding eigenvalue is close to zero. Euclidean distance classifier is adopted in recognition. Two standard databases from Yale University and Olivetti research laboratory are selected to evaluate the recognition accuracy of the proposed method. These databases include face images with different expressions, small occlusion, different illumination condition and different poses. Experimental results on the two face databases show the effectiveness of the method and its insensitivity to the face expression, illumination and posture.© (2004) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

As a result of the analysis of the literature, the based methods of face recognition on fragments of color images were identified. These are flexible comparison in graphs, hidden Markov models, principal component analysis, and neural network methods. The analyzed methods of face recognition are mainly characterized by significant computational costs and low recognition performance. An exception is the neural network methods of face recognition, which, after completing the training, make it possible to obtain a high recognition performance at low computational costs. However, when changing the prototype images of faces, it often becomes necessary to redefine the network architecture and retrain the network. The specificity of neural network methods is also the complexity of selecting the network architecture and its training. Such papers are devoted to the use of neural networks onlyfor extraction of feature vectors of face images. The classification of the obtained feature vectors is then performed by knownmethods, namely, thresholding, a linear support vector machine, nearest neighbors, random forest. It has been observed that the lighting conditions in which the images were obtained and the turning of the head affect the shape of the separating surface and can decrease the feature vector classification performance for face images. Therefore, to improve the classification performance, it was decided to use correlation for prototype matching, a non-linear support vector machine and logistic regression. The performed experiment showed that correlation for prototype matching of low-light face images is characterized by higher classification performance compared to the thresholding. Moreover, the use of the Pearson and Spearman correlation coefficients showed similar results, and when using the Kendall correlation coefficient, the worst classification performance was obtained compared to the Pearson andSpearman coefficients. The research of the classification performance of images of faces that differ in head turn using correlation for prototype matching, a non-linear support vector machine and logistic regression showed the following. Correlation for prototype matching is more appropriate to use with small amounts of data due to the high classification performance and low computational complexity, since a small amount of data does not require a significant number of comparisons. However, on large amounts of data, the non-linear support vector machine requires less computation and shows similar classification performance. Using the results of the experiment, the researcher can select classification methods for a specific set of face images, preliminarily representing them with feature vectors using the network FaceNet.

Aiming to the issue of face recognition with partial contiguous occlusion, a new face recognition method was proposed by removing the outlier area in this paper. A mean face image is firstly obtained from train images, which is subtracted by the test face to form an error face image. Then the error face image is used to obtain the occlusion area of the test image by image segmentation technique, and the train images and test image are tailored by removing the corresponding occlusion area. Finally, face recognition is performed by linear regression classifier or sparse coding classifier. Compared to the similar works, the proposed method has considerably recognition performance improvement with relatively simple computational complexity. Simulation experimental results based on the standard AR face database show effectiveness of this proposed method.

In order to recognize faces, face recognition methods need to be studied. When a face is identified by the current method, the image denoising effect is poor, the face image recognition result thus has error, the time used to recognize the face image is long, the signal to noise ratio, the recognition result and the recognition efficiency are low. Based on the GA-BP neural network algorithm, a face recognition method is proposed. A mixed denoising model of face images is constructed by combining dictionary based sparse representation with non-local similarity. The principal component analysis method is used to extract the feature of the face image after denoising and staining the eigenvector of the face image. The GA-BP neural network algorithm is used to optimize the initial weights and thresholds so as to achieve the optimal value. The feature vectors of face images are ted into the genetic neural network to complete face recognition. Experimental results show that the proposed method has high signal-to-noise ratio, accuracy and recognition efficiency.

Equations for calculation of the constants of biparametrical types of enzyme inhibition and activation were obtained that take into account a ratio of the lengths of L vector projections representing such reactions in the three-dimensional K (m)V I coordinate system. This allows higher accuracy of calculation and is more correct for comparison of these constants. Examples of data analysis of enzyme inhibition and activation by using the traditional equations (they do not take into account the lengths of vector projections) and corrected ones (they take into account the lengths of vector projections) are given. The corrected and traditional equations are used for calculation of the constants of biparametrical types of enzyme inhibition and activation.

A new character recognition method using multiple subspace for each category is proposed. The method is essentially the CLAFIC method. Five subspaces for each category are spanned by vector series expansions constructed from five feature vectors extracted from character patterns separately. The features used are proposed for the pre-classification. The decision procedure of this method is as follows. The feature vectors are extracted from an input character pattern. The projections of those feature vectors on corresponding feature subspaces are computed and five projection lengths are obtained for each category. Five-dimensional vectors whose elements are the above projection lengths are defined for each category. The decision rule is to classify the input pattern into the category on whose five-dimensional vector it has the largest Euclidean norm. The discrimination ability for similar Kanji characters of this method is examined through a discrimination experiment using handprinted Kanji character database ETL-9(B). >

In this paper, a pseudo training sample method is proposed for face recognition using 1D HMMs. The proposed approach uses KL transformed vectors of face images to train ergodic HMMs and later for recognition. In the training process, besides the real face images, pseudo face images are generated using the real images and also used for training. Recognition accuracy is improved a lot in this way and analysis is given for the improvement. Our method achieves good recognition accuracy on ORL face database. The pseudo training sample method is very useful for most practical situations that there are not enough images for training.

In multivariate chaotic time series prediction, correlation analysis is important for reducing input dimensions and improving prediction performance. Grey relational analysis (GRA) has proved to be an effective method for data correlation analysis, especially for inexact data and incomplete data. In GRA, points are usually regarded as objects, and the distance between points or the concave and convex degree are mostly used to measure the correlations. However, with discrete variables, correlation analysis results always tend to have some deviations when using prior GRA methods. Furthermore, GRA methods cannot directly use vector datasets. Therefore, in this paper, an improved GRA method is proposed based on vector projections. The input and output variables are expressed as vectors by linking two adjacent points. The vectors, instants of the points, are regarded as the objects, and the projection length of input variables to output variables is used to measure the correlations. The smaller the difference between the projection length and the input variables, the higher the correlation. Then, a hybrid variable selection and prediction model is proposed based on the improved GRA method for multivariate chaotic time series predictions, in order to overcome the negative effects of irrelevant and redundant variables caused by phase-space reconstruction. The experimental results based on the gas furnace dataset and San Francisco river runoff dataset demonstrate that the improved GRA method is effective for data correlation analysis, and the prediction accuracy is better than prior GRA-based methods.

Background subtraction is a useful and effective method for detecting moving objects in video images. Since this method assumes that image variations are caused only by moving objects (i.e., the background scene is assumed to be stationary), however, its applicability is limited. In this paper, we propose a background subtraction method that robustly handles various changes in the background. The method learns the chronological changes in the observed scene's background in terms of distributions of image vectors. The method operates the subtraction by evaluating the Mahalanobis distances between the averages of such image vectors and newly observed image vectors. The method we propose herein expresses actual changes in the background using a multi-dimensional image vector space. This enables the method to detect objects with the correct sensitivity. We also introduce an eigenspace to reduce the computational cost. We describe herein how approximate Mahalanobis distances are obtained in this eigenspace. In our experiments, we confirmed the proposed method's effectiveness for real world scenes.