Face recognition under large illumination variations using homomorphic filtering in spatial domain

Performance of a contemporary two-dimensional face-recognition system has not been satisfied due to the variation in lighting. As a result, many works of solving illumination variation in face recognition have been carried out in past decades. Among them, the Illumination-Reflectance model is one of the generic models that is used to separate the individual reflectance and illumination components of an object. The illumination component can be removed by means of image-processing techniques to regain an intrinsic face feature, which is depicted by the reflectance component. We present a wavelet-based illumination invariant algorithm as a preprocessing technique for face recognition. On the basis of the multiresolution nature of wavelet analysis, we decompose both illumination and reflectance components from a face image in a systematic way. The illumination component wherein resides in the low-spatial-frequency subband can be eliminated efficiently. This technique works out very advantageously for achieving higher recognition performance on YaleB, CMU PIE, and FRGC face databases.

A comparative study on local binary pattern (LBP) based face recognition: LBP histogram versus LBP image

A comprehensive account of almost all the feature extraction methods used in current face recognition systems is presented. Specifically we have made distinction in the holistic and local feature extraction and differentiate them qualitatively as opposed to quantitatively. It is argued that a global feature representation should be preferred over a bag-of-feature approach. The problems in current feature extraction techniques and their reliance on a strict alignment is discussed. Finally we have introduced to use face-GLOH signatures that are invariant with respect to scale, translation and rotation and therefore do not require properly aligned images. The resulting dimensionality of the vector is also low as compared to other commonly used local features such as Gabor, LBP etc. and therefore learning based methods can also benefit from it. In a typical multi-view face recognition task, where it is assumed to have several examples of a subject available for training, we have shown in an extensive experimental setting the advantages and weaknesses of commonly used feature descriptions. Our results show that under more realistic assumptions, most of the classifiers failed on conventional features. While using the introduced face-GLOH-signature representation is relatively less affected by large in-class variations. This has been demonstrated by providing a fair performance comparison of several classifiers under more practical conditions such as misalignments, large number of subjects and large pose variations. An important conclusion is to be drawn from the results on FERET is that conventional multi-view face recognition cannot cope well with regards to large pose variations. Even using a large number of training examples in different poses for a subject do not suffice for a satisfactory recognition. In order to solve the problem where only one training example per subject is available, many recent methods propose to use image synthesis to generate a given subject at all other views and then perform a conventional multi-view recognition (Beymer and Poggio, 1995; Gross et al, 2004). Besides the fact that such synthesis techniques cause severe artefacts and thus cannot preserve the identity of an individual, a conventional classification cannot yield good recognition results, as has been shown in an extensive experimental setting. More sophisticated methods are therefore needed in order to address pose invariant face recognition. Large pose differences cause significant appearance variations that in general are larger than the appearance variation due to identity. One possible way of addressing this is to learn these variations across each pose, more specifically by fixing the pose and establishing a correspondence on how a person’s appearance changes under this pose one could reduce the in-class appearance variation significantly. In our very recent work (Sarfraz and Hellwich, 2009), we demonstrate the usefulness of face-GLOH signature in this direction.

This paper proposes a novel face representation approach, local Gabor binary mapping pattern (LGBMP), for multi-view gender classification. In this approach, a face image is first represented as a series of Gabor magnitude pictures (GMP) by applying multi-scale and multi-orientation Gabor filters. Each GMP is then encoded as a LGBP image where a uniform local binary pattern (LBP) operator is used. After that, each LGBP image is divided into non-overlapping rectangular regions, from which spatial histograms are extracted. Although an LGBP feature vector can be obtained by fitting together the regional histograms, it can not be employed in pattern classification due to its high dimension. We propose that each regional LGBP feature be mapped onto a one-dimensional subspace independently before they are concatenated as a whole feature vector. This is attractive since we reduce the feature dimension and also preserve the spatial information of LGBP image. Two ways have been proposed to map the regional LGBP feature in this paper. One is so-called LGBMP-LDA using linear discriminant analysis (LDA) for dimensionality reduction while the other is to project the regional LGBP feature onto the class center connecting line, namely, LGBMP-CCL. As a result, despite several decades of Gabor filters, the final feature dimension is even less than that of the feature extracted by using LBP directly on gray-scale images. The classification tasks in our work are performed by support vector machines (SVM). The experimental results on the CAS-PEAL face database indicate that the proposed approach achieves higher accuracy than the others such as SVMs+Gray-scale pixel, SVMs+Gabor and SVMs+LBP approach, more particularly, it has the lowest dimension of feature vector.

Visible face recognition systems, being vulnerable to illumination, expression, and pose, can not achieve robust performance in unconstrained situations. Meanwhile, near infrared face images, being light- independent, can avoid or limit the drawbacks of face recognition in visible light, but its main challenges are low resolution and signal noise ratio (SNR). Therefore, near infrared and visible fusion face recognition has become an important direction in the field of unconstrained face recognition research. In this paper, a novel fusion algorithm in non-subsampled contourlet transform (NSCT) domain is proposed for Infrared and visible face fusion recognition. Firstly, NSCT is used respectively to process the infrared and visible face images, which exploits the image information at multiple scales, orientations, and frequency bands. Then, to exploit the effective discriminant feature and balance the power of high-low frequency band of NSCT coefficients, the local Gabor binary pattern (LGBP) and Local Binary Pattern (LBP) are applied respectively in different frequency parts to obtain the robust representation of infrared and visible face images. Finally, the score-level fusion is used to fuse the all the features for final classification. The visible and near infrared face recognition is tested on HITSZ Lab2 visible and near infrared face database. Experiments results show that the proposed method extracts the complementary features of near-infrared and visible-light images and improves the robustness of unconstrained face recognition.

An efficient face recognition system should recognize faces in different views and poses. The efficiency of a human face recognition system depends on the capability of face recognition in presence of changes in the appearance of face due to expression, pose and illumination. A novel algorithm which utilizes the combination of texture and depth information based on Modular PCA to overcome the problem of pose variation and illumination change for face recognition is proposed. The system has combined 2D and 3D systems in the feature level which presents higher performance in contrast with methods which utilizes either 2D or 3D system separately. A multimodal face recognition based on Modular PCA when compared with conventional PCA algorithm has an improved recognition rate for face images with large variations in illumination and facial expression. The proposed algorithm is tested with FRAV3D database that has faces with pose variation and illumination changes. Recognition rates from experimental results show the superiority of Modular PCA over conventional PCA methods in tackling face images with different pose variations and changes in illuminations. The proposed algorithm shows a recognition rate of 86% that is achieved in fusion experiment.

Our objective was to evaluate the clinical feasibility of spatial domain filtering as an alternative to additional image reconstruction using different kernels in chest CT. Spatial domain filtering generates smooth images from sharp images and thus avoids the need for additional reconstructions when two sets of images are desired. Forty adult patients with clinical suspicion of pulmonary embolism were examined utilizing multi-slice CT (Somatom Volume Zoom, Siemens, Germany). Derived from thin collimated source images (100 mAs, collimation 4x1 mm, rotation time 0.5 s, table speed 7 mm/rotation), two sets of images [effective slice thickness (S(eff)) 5 mm, reconstruction increment (RI) 5 mm) were generated using lung (Siemens B50) and soft tissue (Siemens B30) kernels. Additionally, B50 images were filtered in the spatial domain, producing images largely equivalent to B30 images. Firstly, diagnostic accuracy was assessed on spatial domain filtered images regarding central, segmental, and subsegmental pulmonary embolism. In a second step, diagnostic accuracy was assessed for the initially reconstructed B30 images. The results were compared with thin axial slices from the same data set, which were considered as the gold standard in this respect (S(eff )1.25 mm, RI 0.8 mm; B30). Initially reconstructed B30 slices and secondary filtered images were rated for subjective image quality, using a five-point scale (1=excellent, 2=good, 3=moderate, 4=poor, 5=non-diagnostic). Finally, quantitative measurements were assessed using the region of interest (ROI) methodology. In 20 patients pulmonary embolism was proven. Five-millimeter images revealed 10 of 10 central emboli, 18 of 19 segmental thrombi, and 18 of 20 emboli on the subsegmental level. Pulmonary embolism was excluded in 18 of 20 subjects, and in 2 patients a false-positive result was obtained in subsegmental arteries. These findings were concordant for reconstructed and filtered images. Quantitative density measurements provided comparable Hounsfield units in this respect. Subjective gradings of image quality, based on soft tissue settings, were 1.30 (+/-0.61) for reconstructed slices vs 1.35 (+/-0.62) for filtered images (weighted kappa coefficient 0.6117; 95% confidence intervals 0.3298-0.8935). Spatial domain filtering has proved to be feasible. Compared with conventional soft tissue reconstructions for central, segmental, and subsegmental pulmonary embolism, no significant difference in the diagnostic value of spatial domain filtered images was found. Online modifications of image sharpness and pixel noise in real time leads to a considerable reduction of processing time and cost saving for storage of CT images. Despite different data processing methods, thin effective slice thicknesses and overlapping reconstruction increments are mandatory for detailed CT analysis of pulmonary embolism on the segmental and subsegmental level.

Illumination variation significantly affects the performance of face recognition algorithms. However, it is difficult to construct a database that meets the demand of illumination analysis with current face image database construction technology. The influence of illumination variation on deep face recognition based on 3D face models is investigated. First a new method is presented to produce face images under arbitrary illuminations by using a 3D face model based on face basis image representation theory proposed. Then the database used for illumination analysis is constructed. Afterwards, an optimal way is explored to select illuminations used for constructing a high quality face image database by analyzing the impacts of different illumination sampling strategies on the performances of face recognition models. With the advantage of accurate illumination labeling of virtual data, the performances of different illumination labeling methods on the training of face recognition network are tested on a multi-task learning framework. Through experiments on the effects of different lighting sampling schemes and labeling methods on the performances of face recognition models carried out on virtual data and real data, it is concluded that using an appropriate amount of lightings of basis images to build a database is an effective lighting sampling scheme, and accurate lighting annotation can help further improve the face recognition rate, the corresponding recognition models have face recognition accuracy rates of more than 98% on virtual and real monitoring datasets which contain face images captured under very challenging lighting conditions. Proposed work improve the performance of deep face recognition models, and facilitate the construction of high quality face image databases with various illuminations.

Biometrics is the study of methods of recognizing humans based on their behavioral and physical characteristics or traits. Face recognition is one of the biometric modalities that received a great amount of attention from many researchers during the past few decades because of its potential applications in a variety of security domains. Face recognition however is not only concerned with recognizing human faces, but also with recognizing faces of non-biological entities or avatars. Fortunately, the need for secure and affordable virtual worlds is attracting the attention of many researchers who seek to find fast, automatic and reliable ways to identify virtual worlds’ avatars. In this work, I propose new techniques for recognizing avatar faces, which also can be applied to recognize human faces. Proposed methods are based mainly on a well-known and efficient local texture descriptor, Local Binary Pattern (LBP). I am applying different versions of LBP such as: Hierarchical Multi-scale Local Binary Patterns and Adaptive Local Binary Pattern with Directional Statistical Features in the wavelet space and discuss the effect of this application on the performance of each LBP version. In addition, I use a new version of LBP called Local Difference Pattern (LDP) with other well-known descriptors and classifiers to differentiate between human and avatar face images. The original LBP achieves high recognition rate if the tested images are pure but its performance gets worse if these images are corrupted by noise. To deal with this problem I propose a new definition to the original LBP in which the LBP descriptor will not threshold all the neighborhood pixel based on the central pixel value. A weight for each pixel in the neighborhood will be computed, a new value for each pixel will be calculated and then using simple statistical operations will be used to compute the new threshold, which will change automatically, based on the pixel’s values. This threshold can be applied with the original LBP or any other version of LBP and can be extended to work with Local Ternary Pattern (LTP) or any version of LTP to produce different versions of LTP for recognizing noisy avatar and human faces images.

In this paper, face recognition technology is studied and explored, and the classical local binary pattern (LBP) algorithm is studied and improved. An improved LBP algorithm is proposed to solve the deficiency of LBP: a "double circle" LBP algorithm is proposed, which improves the rotation invariability of LBP to a greater extent and has a good recognition effect on small-pose face images. On the basis of good recognition effect of LBP based on partitioning, the idea of "multiple partitioning + intermediate partitioning" is put forward, which effectively makes up for the shortage of incomplete extraction of peripheral information of original segmentation line. Facial features are the key features of face recognition. Different organs contribute differently to face image recognition. The idea of "key block" based on the original block is proposed. Finally, the application of an LBP algorithm for face recognition is taken as an example, and the experimental results verify the reliability of the face recognition method based on LBP for weak light and expression change images.

Face recognition is a common means of identity authentication. Mobile learning platform login technology has developed from user name and password to face recognition. In order to improve effectively the rate of face recognition, this paper proposes a kind of face recognition algorithm based on self-adaptive blocking local binary pattern (LBP) and dual channel convolutional neural network (CNN) with different convolution kernels. Firstly, the Gamma correction, the Mallet wavelet filtering and normalization are used to preprocess the face image. The face image is decomposed and reconstructed by 2-layer Mallet wavelet to filter out the interference signal effectively. Although the general LBP operator extracts the overall texture and contour features of the face image, the distribution of the bright spot, dark spot and other micro details cannot be fully characterized. In order to solve this problem, integral projection is introduced to project the image horizontally and vertically. The extreme points of the projection represent the texture mutation points of the face image in the horizontal and vertical directions. These extreme points are used as the boundary of the image blocking, and the LBP value of the face image is extracted by the self-adaptive blocking strategy. Combining the features of k-nearest neighbor classifier and softmax, a k-softmax classification method is proposed to classify and recognize the face image labels. After two channel network structure training, this method is tested on Yale, ORL, extended Yale B and self-built face databases by five experiments, comparing with other face recognition algorithms. The results show that the proposed method based on SAB-LBP and dual channel CNN has high recognition rate and computational efficiency.

<p>Various face recognition methods are derived using local features among them the Local Binary Pattern (LBP) approach is very famous. The histogram techniques based on LBP is a complex task. Later Uniform Local Binary Pattern (ULBP) is derived on LBP, based on the bitwise transitions and ULBP’s are treated as the fundamental property of texture. The ULBP approach treated all Non-Uniform Local Binary Patterns’ (NULBP) into one miscellaneous label. Recently we have derived Prominent LBP (PLBP), Maximum PLBP (MPLBP) and Smallest PLBP (SPLBP). The PLBP consists of the majority of the ULBP’s and some of the NULBP’s. The basic disadvantage of these various variants of LBP’s is they are basically local approaches and completely failed in representing features derived from large regions or macrostructures, which are very much essential for faces. This paper derives PLBP’s on the large region. The rectangular region of this paper is assumed with a size of multiples of three and PLBPs are evaluated on dividing each region into multiple regions. The proposed Multi Region-PLBP (MR-PLBP) approach is tested on three facial databases namely Yale, Indian and AT&T ORL. The experimental results show the proposed approach significantly outperforms the other LBP based face recognition methods.</p>

Call for Papers: Special Issue ofComputer Vision and Image Understandingon the Analysis of Volumetric Images

Face recognition is an area of computer vision that has attracted a lot of interest from the research community. A growing demand for robust face recognition software in security applications has driven the development of interesting approaches in this field. A large quantity of research in face recognition deals with visible face images. In the visible spectrum the illumination and face expressions changes represent a significant challenge for the recognition system. To avoid these problems, researchers proposed recently the use of 3D and infrared imaging for face recognition. In this work, we introduce a new framework for infrared face recognition using texture descriptors. This framework exploits linear and non linear dimensionality reduction techniques for face learning and recognition in the texture space. Active and passive infrared imaging modalities are used and comparison with visible face recognition is performed. Two multispectral face recognition databases were used in our experiments: Equinox Database (Visible, SWIR, MWIR, LWIR) and Laval University Multispectral Database (Visible, NIR, MWIR, LWIR). The obtained results show high increase in recognition performance when texture descriptors like LBP (Local Binary Pattern) and LTP (Local Ternary Pattern) are used. The best result was obtained in the short wave infrared spectrum (SWIR) using non linear dimensionality reduction techniques.

Serving as an effective texture description operator, local binary pattern (LBP) has been applied in the visible face recognition successfully. In order to enhance the recognition performance, the technology of face recognition depending on the near infrared (NIR) image has attracted extensive attention in the recent years. Although the characteristics of LBP have been researched in optical image thoroughly, the development of these do not catch enough attention in the NIR image at all. Therefore, in our paper, we study the characteristics of LBP from the following two aspects in NIR image. On one hand, we come to the probability distribution of various patterns of LBP in the NIR facial image. On the other hand, we discuss the influence to LBP caused by the illumination change in facial image.