Face detection is one of the most challenging and long-studied areas in computer vision. In real-world, images are exposed to the noise and degradation. In this paper, we investigate the robustness of two networks namely SSD and Faster R-CNN in confrontation with salt and pepper noise, Gaussian blur, as well as JPEG compression. Our experiments are conducted on the well-known Wider Face dataset. These experiments show that the Faster R-CNN is more robust against Gaussian blur, while SSD is much more sensitive to the edges. On the other hand, SSD is more robust against reduced-quality JPEG compressed images. The reason should be due to the sensitivity of Faster R-CNN to the texture of the objects. Moreover, our experiments demonstrated that both networks have a relatively similar resistance under salt and pepper noise.

In biometric recognition, feature fusion is an important area of research due to the fact that multiple types of features contain richer and complementary information. Discriminative Correlation Analysis (DCA) is a recently proposed feature fusion method, which incorporates the class association into correlation analysis so that the features not only have the maximum intrinsic correlation between feature sets but also have class structure information. However, DCA is a linear technique, that finds a linear transformation of the original space. For highly nonlinearly distributed data, classification with nonlinear techniques works better than the linear ones. In this paper, we propose Kernel-DCA which generalizes DCA in order to handle nonlinear problems. Similar to Kernel-SVM, Kernel-DCA utilizes a kernel method to map feature sets to a high-dimensional space in which features are linearly separable. Experimental results, for the fusion of ear and face feature, using the WVU database with large variations in pose, show that Kernel-DCA achieves better results on nonlinearly distributed data than DCA and other feature fusion methods.

Due to large distances between surveillance cameras and subjects, the captured images usually have low resolution in addition to uncontrolled poses and illumination conditions that adversely affect the performance of face recognition algorithms. In this paper, we present a low-resolution face recognition technique based on Discriminant Correlation Analysis (DCA). DCA analyzes the correlation of the features in high-resolution and low-resolution images and aims to find projections that maximize the pair-wise correlations between the two feature sets and at the same time, separate the classes within each set. This makes it possible to project the features extracted from high-resolution and low-resolution images into a common space, in which we can apply matching. The proposed method is computationally efficient and can be applied to challenging real-time applications such as recognition of several faces appearing in a crowded frame of a surveillance video. Extensive experiments performed on low-resolution surveillance images from the SCface database as well as FRGC database demonstrated the efficacy of our proposed approach in the recognition of low-resolution face images, which outperformed other state-of-the-art techniques.

Information fusion is a key step in multimodal biometric systems. The fusion of information can occur at different levels of a recognition system, i.e., at the feature level, matching-score level, or decision level. However, feature level fusion is believed to be more effective owing to the fact that a feature set contains richer information about the input biometric data than the matching score or the output decision of a classifier. The goal of feature fusion for recognition is to combine relevant information from two or more feature vectors into a single one with more discriminative power than any of the input feature vectors. In pattern recognition problems, we are also interested in separating the classes. In this paper, we present discriminant correlation analysis (DCA), a feature level fusion technique that incorporates the class associations into the correlation analysis of the feature sets. DCA performs an effective feature fusion by maximizing the pairwise correlations across the two feature sets and, at the same time, eliminating the between-class correlations and restricting the correlations to be within the classes. Our proposed method can be used in pattern recognition applications for fusing the features extracted from multiple modalities or combining different feature vectors extracted from a single modality. It is noteworthy that DCA is the first technique that considers class structure in feature fusion. Moreover, it has a very low computational complexity and it can be employed in real-time applications. Multiple sets of experiments performed on various biometric databases and using different feature extraction techniques, show the effectiveness of our proposed method, which outperforms other state-of-the-art approaches.

In this paper, we present Discriminant Correlation Analysis (DCA), a feature level fusion technique that incorporates the class associations in correlation analysis of the feature sets. DCA performs an effective feature fusion by maximizing the pair-wise correlations across the two feature sets, and at the same time, eliminating the between-class correlations and restricting the correlations to be within classes. Our proposed method can be used in pattern recognition applications for fusing features extracted from multiple modalities or combining different feature vectors extracted from a single modality. It is noteworthy that DCA is the first technique that considers class structure in feature fusion. Moreover, it has a very low computational complexity and it can be employed in realtime applications. Multiple sets of experiments performed on various biometric databases show the effectiveness of our proposed method, which outperforms other state-of-the-art approaches.

Single sample face recognition have become an important problem because of the limitations on the availability of gallery images. In many real-world applications such as passport or driver license identification, there is only a single facial image per subject available. The variations between the single gallery face image and the probe face images, captured in unconstrained environments, make the single sample face recognition even more difficult. In this paper, we present a fully automatic face recognition system robust to most common face variations in unconstrained environments. Our proposed system is capable of recognizing faces from non-frontal views and under different illumination conditions using only a single gallery sample for each subject. It normalizes the face images for both in-plane and out-of-plane pose variations using an enhanced technique based on active appearance models (AAMs). We improve the performance of AAM fitting, not only by training it with in-the-wild images and using a powerful optimization technique, but also by initializing the MM with estimates of the locations of the facial landmarks obtained by a method based on flexible mixture of parts. The proposed initialization technique results in significant improvement of AAM fitting to non-frontal poses and makes the normalization process robust, fast and reliable. Owing to the proper alignment of the face images, made possible by this approach, we can use local feature descriptors, such as Histograms of Oriented Gradients (HOG), for matching. The use of HOG features makes the system robust against illumination variations. In order to improve the discriminating information content of the feature vectors, we also extract Gabor features from the normalized face images and fuse them with HOG features using Canonical Correlation Analysis (CCA). Experimental results performed on various databases outperform the state-of-the-art methods and show the effectiveness of our proposed method in normalization and recognition of face images obtained in unconstrained environments. (C) 2015 Elsevier Ltd. All rights reserved.

In this paper, we present methods for segmenting noisy two-dimensional forward-scan sonar images and classify and model their background. The segmentation approach differentiates the highlight blobs, cast shadows, and the background of sonar images. There is usually little information within relatively large background regions corresponding to the flat sea bottom and (or) water column, as they are often corrupted with speckle noise. Our experiments show that the background texture is dominated by the speckle noise which has the appearance of a pseudo-random texture. We show that the background texture of the underwater sonar images can be categorized by a small number of classes. The statistical features work better than the texture-based features in categorizing the pseudo-random background, which further strengthen our hypothesis of the dominance of noise over the background texture. As a result, we can model the noisy background with a few parameters. This has an application in coding the sonar images in which highlight blob regions and cast shadows are coded at the encoder side while the speckle noise-corrupted background can be synthesized at the decoder side. Since the background regions occupy a large fraction of the FS sonar image, we expect higher compression rates than most current image or video coding standards and other custom-designed sonar image compression techniques.

In biometric identification systems, the biometric database is typically stored in a trusted server, which is also responsible for performing the identification process. However, a standalone server may not be able to provide enough storage and processing power for large databases. Nowadays, cloud computing and storage solutions have provided users and enterprises with various capabilities to store and process their data in third-party data centers. However, maintenance of the confidentiality and integrity of sensitive data requires trustworthy solutions for storage and processing of data with proven zero information leakage. In this paper, we present CloudID, a privacy-preserving cloud-based and cross-enterprise biometric identification solution. It links the confidential information of the users to their biometrics and stores it in an encrypted fashion. Making use of a searchable encryption technique, biometric identification is performed in encrypted domain to make sure that the cloud provider or potential attackers do not gain access to any sensitive data or even the contents of the individual queries. In order to create encrypted search queries, we propose a k-d tree structure in the core of the searchable encryption. This helps not only in handling the biometrics variations in encrypted domain, but also in improving the overall performance of the system. Our proposed approach is the first cloud-based biometric identification system with a proven zero data disclosure possibility. It allows different enterprises to perform biometric identification on a single database without revealing any sensitive information. Our experimental results show that CloudID performs the identification of clients with high accuracy and minimal overhead and proven zero data disclosure. (C) 2015 Elsevier Ltd. All rights reserved.

Objective image quality assessment has always been a challenging issue among image processing scientists. Here, a novel image quality metric based on mutual information of image features is presented. The proposed metric calculates the amount of information that one image contains about the other. Normalized feature vectors extracted from the test and reference images represent the considered information as the marginal probability distributions. In order to use the mutual information we need not only the marginal probability distributions but also the joint distribution between two images. In this work, Nelsen’s method is employed to estimate the joint probability distribution from marginal distributions using the correlation between two feature vectors. Experimental results and comparisons with other well-known full-reference metrics like SSIM, MSSIM, FSIM and SFF certify the soundness of our proposed metric.

In this paper, we present a non-reference image fusion metric based on the mutual information of image features. Whereas a recent metric proposed by the author called FMI achieves such a goal, the algorithm is complex and has high memory requirements for its calculations. This paper shows how to modify the model of FMI, and proposes a faster algorithm to achieve similar results. The new algorithm achieves a significant complexity reduction in comparison to the previous model. Various experiments prove the efficiency of the algorithm in consistency with the subjective criteria. Matlab source code for this metric is provided at http://www.mathworks.com/matlabcentral/fileexchange/45926.

In this paper, we present a computationally efficient statistical face modeling approach. The efficiency of our proposed approach is the result of mathematical simplifications in the core formula of a previous face modeling method and the use of the singular value decomposition. In order to reduce the errors in our resulting models, we preprocess the facial images to normalize for pose and illumination and remove little occlusions. Then, the statistical face models for the enrolled subjects are obtained from the normalized face images. The effects of the variations in pose, facial expression, and illumination on the accuracy of the system are studied. Experimental results demonstrate the reduction in the computational complexity of the new approach and its efficacy in modeling the face images.

Biometric identification is a challenging subject among computer vision scientists. The idea of substituting biometrics for passwords has become more attractive after powerful identification algorithms have emerged. However, in this regard, the confidentiality of the biometric data becomes of a serious concern. Biometric data needs to be securely stored and processed to guarantee that the user privacy and confidentiality is preserved. In this paper, a method for biometric identification using encrypted biometrics is presented, where a method of search over encrypted data is applied to manage the identification. Our experiments of facial identification demonstrate the effective performance of the system with a proven zero information leakage.

The ultimate goal of pattern recognition is to discriminate different classes with minimum misclassification rate. The feature vector used in classification should be as short as possible to reduce the algorithm complexity and informative enough to be able to discriminate complicated patterns. In this regard, dimensionality reduction methods are utilized to reduce the raw feature vector length and also to make the features more discriminative. In this paper, a face detection scheme is proposed by using discrete cosine transform (DCT) features in Bayesian discriminating features (BDF) classifier. Low redundancy of DCT features, optimal reconstruction property of Hotelling transform as the dimensionality reduction method, and the minimum error rate of Bayesian classifier, all in all, bring about a high detection rate in the proposed scheme. Various experiments, performed on different databases, certify that using more informative feature vectors results in a higher dimensionality reduction and improves the classifier's detection rate.

Principal component analysis (PCA) is an effective tool for dimension reduction in classification approaches. Bayesian discriminating features (BDF) is a classifier which effectively utilizes this tool. In this classifier, any of the M largest eigenvalues of the training patterns' covariance matrix are individually involved in classification while the arithmetic average of the remaining eigenvalues take part just as a single parameter. In this paper, by suggesting a new classifier, effect of the number of involved eigenvalues in classification performance is studied. In the suggested classifier we ignore the arithmetic average that is utilized in BDF. Our experiments verify that increasing M does not lead to an ongoing increase in classifier's detection rate in both BDF and the proposed one. However, by over-increasing M, the dependency of classifiers' parameters to the training samples increases which could reduce the performance of the classifiers when they come to make decision about new samples. Furthermore, experimental results verify that arithmetic average of the remaining eigenvalues in BDF improves the classifier performance only when an appropriate number of eigenvalues is selected; hence, ignoring the arithmetic average, as done in proposed classifier, could provide a better performance rather than BDF.

The objective of image fusion is to combine relevant information from multiple images into a single image. The discrete cosine transform (DCT) based methods of image fusion are more efficient and time-saving in real-time systems using DCT based standards of still image or video. Existing DCT based methods are suffering from some undesirable side effects like blurring or blocking artifacts which reduce the quality of the output image. Furthermore, some of these methods are rather complex and this contradicts the concept of the simplicity of DCT based algorithms. In this paper, an efficient approach for fusion of multi-focus images based on variance calculated in DCT domain is presented. Due to simplicity of our proposed method, it can be easily used in real-time applications. The experimental results verify the efficiency improvement of our method both in output quality and complexity reduction in comparison with several recent proposed techniques.

The widespread usage of image fusion causes an increase in the importance of assessing the performance of different fusion algorithms. The problem of introducing a suitable quality measure for image fusion lies in the difficulty of defining an ideal fused image. In this paper, we propose a non-reference objective image fusion metric based on mutual information which calculates the amount of information conducted from the source images to the fused image. The considered information is represented by image features like gradients or edges, which are often in the form of two-dimensional signals. In this paper, a method of estimating the joint probability distribution from marginal distributions is also presented which is employed in calculation of mutual information. The proposed method is compared with the most popular existing algorithms. Various experiments, performed on several databases, certify the efficiency of our proposed method which is more consistent with the subjective criteria.

The objective of image fusion is to combine information from multiple images of the same scene in order to deliver only the useful information. The discrete cosine transform (DCT) based methods of image fusion are more suitable and time-saving in real-time systems using DCT based standards of still image or video. In this paper an efficient approach for fusion of multi-focus images based on variance calculated in DCT domain is presented. The experimental results on several images show the efficiency improvement of our method both in quality and complexity reduction in comparison with several recent proposed techniques.