International Journal of Pattern Recognition and Artificial Intelligence

Low rank representation (LRR) is widely used to construct a good affinity matrix to cluster data drawn from the union of multiple linear subspaces. However, it is not easy to solve the LRR problem in a closed form, and augmented Lagrange multiplier method (ALM) is usually applied. ALM takes a relative long time dealing with the real-world data. To solve the LRR problem efficiently, we propose an efficient low rank representation (eLRR) algorithm. Given a contaminated data set, we propose a novel way to solve the LRR of the data. We establish a useful theorem which directly gives an approximate solution to our LRR optimization problem. Thus, we can construct a good affinity matrix for subspace clustering. Experimental results with several public databases verify the efficiency and effectiveness of our method.

In this paper a platform is presented, that given a stochastic context-free grammar (SCFG), automatically outputs the description of the parser in synthesizable hardware description language (HDL) which can be downloaded in an Field Programmable Gate Arrays (FPGA) board. Initially, according to our methodology the SCFG is augmented with attributes which store the probability values and can be evaluated through corresponding stack actions. The architecture of the produced system is based on a proposed extension of Earley’s parallel algorithm, which given an input string, generates the parse trees in the form of an AND-Or parse tree. This AND-or parse tree is then traversed using a proposed tree traversal technique in order to execute the corresponding actions in the correct order, so as to compute the necessary probabilities. The platform is suitable for embedded systems applications where a natural language interface is required or in pattern recognition tasks. The parser generated by the presented platform has been tested for various SCFGs and compared to software approaches. The performance comparison is one to two orders of magnitude in favor of the presented hardware, compared to previous software approaches, depending on the application, the input string length and the number of produced trees.

The interplay of machine learning (ML) and optimization methods is an emerging field of artificial intelligence. Both ML and optimization are concerned with modeling of systems related to real-world problems. Parameter selection for classification models is an important task for ML algorithms. In statistical learning theory, cross-validation (CV) which is the most well-known model selection method can be very time consuming for large data sets. One of the recent model selection techniques developed for support vector machines (SVMs) is based on the observed test point margins. In this study, observed margin strategy is integrated into our novel infinite kernel learning (IKL) algorithm together with multi-local procedure (MLP) which is an optimization technique to find global solution. The experimental results show improvements in accuracy and speed when comparing with multiple kernel learning (MKL) and semi-infinite linear programming (SILP) with CV.

Digital palaeography is an emerging research area which aims to introduce digital image processing techniques into palaeographic analysis for the purpose of providing objective quantitative measurements. This paper explores the use of a fully automated handwriting feature extraction, visualization, and analysis system for digital palaeography which bridges the gap between traditional and digital palaeography in terms of the deployment of feature extraction techniques and handwriting metrics. We propose the application of a set of features, more closely related to conventional palaeographic assesment metrics than those commonly adopted in automatic writer identification. These features are emprically tested on two datasets in order to assess their effectiveness for automatic writer identification and aid attribution of individual handwriting characteristics in historical manuscripts. Finally, we introduce tools to support visualization of the extracted features in a comparative way, showing how they can best be exploited in the implementation of a content-based image retrieval (CBIR) system for digital archiving.

This paper presents an improved genetic algorithm for the optimization of the structure element used in morphological open and closed filtering. Considering that the evaluation of the froth images of coal flotation is categorized as a no-reference image evaluation, in the optimization of the structural element, a denoising evaluation index of an improved information capacity was used as the adaptation degree function. In addition, this paper proposes the determination method of chromosome length in the structure element optimization algorithm. In the improved genetic algorithm, based on adaptive variation, the variation regulation factor and the mechanism of concentration adjustment are introduced. When compared to an optimization process of the structural element in froth image denoising using the genetic algorithm, the adaptive genetic algorithm, the improved genetic algorithm improves the efficiency and accuracy of the optimization process. It has been proven that optimizing the structural element by the improved genetic algorithm increases fitness and reduces noise when using morphological open and closed reconstruction filters.

Locating the center of the pupils is the most important foundation and the core component of gaze tracking. The accuracy of gaze tracking largely depends on the quality of images, but additional constraints and large amount of calculation make gaze tracking impractical on high-resolution images. Although some eye-gaze trackers can get accurate result, improving the accuracy of pupil feature on low-resolution images and accurately recognizing closed eye images are still common tasks in the field of gaze estimation. Our aim is to get the accurate localization of pupil center on low-resolution image. To this aim, we proposed a simple but effective method which can accurately locate pupil center in real time. The method first gets initial eye center based on improved scale-invariant feature transform (SIFT) descriptor and support vector machine (SVM) classifier, and then gets final position of the pupil center through a size variable correction rectangular block. In this paper, comparing with the reported state-of-the-art methods,the experimental results demonstrate that our system can achieve a more accurate result on low-resolution images. On top of that, our approach shows robustness on closed eye images while some other methods would not recognize the closed eye images.

There are multiple challenges in face detection, including illumination conditions and diverse poses of the user. Prior works tend to detect faces by segmentation at pixel level, which are generally not computationally efficient. When people are sitting in the car, which can be regarded as single face situations, most face detectors fail to detect faces under various poses and illumination conditions. In this paper, we propose a simple but efficient approach for single face detection. We train a deep learning model that reconstructs face directly from input image by removing background and synthesizing 3D data for only the face region. We apply the proposed model to two public 3D face datasets, and obtain significant improvements in false rejection rate (FRR) of 4.6% (from 4.6% to 0.0%) and 21.7% (from 30.2% to 8.5%), respectively, compared with state-of-art performances in two datasets. Furthermore, we show that our reconstruction approach can be applied using 1/2 the time of a widely used real-time face detector. These results demonstrate that the proposed Reconstruction ConNet (RN) is both more accurate and efficient for real-time face detection than prior works.

Recently, considerable research efforts have been devoted to effective utilization of facial color information for improved recognition performance. Of all color-based face recognition (FR) methods, the most widely used approach is a color FR method using input-level fusion. In this method, augmented input vectors of the color images are first generated by concatenating different color components (including both luminance and chrominance information) by column order at the input level and feature subspace is then trained with a set of augmented input vectors. However, in practical applications, a testing image could be captured as a grayscale image, rather than as a color image, mainly caused by different, heterogeneous image acquisition environment. A grayscale testing image causes so-called dimensionality mismatch between the trained feature subspace and testing input vector. Disparity in dimensionality negatively impacts the reliable FR performance and even imposes a significant restriction on carrying out FR operations in practical color FR systems. To resolve the dimensionality mismatch, we propose a novel approach to estimate new feature subspace, suitable for recognizing a grayscale testing image. In particular, new feature subspace is estimated from a given feature subspace created using color training images. The effectiveness of proposed solution has been successfully tested on four public face databases (DBs) such as CMU, FERET, XM2VTSDB, and ORL DBs. Extensive and comparative experiments showed that the proposed solution works well for resolving dimensionality mismatch of importance in real-life color FR systems.

Due to the rapid rise of telemedicine, a lot of patients’ information will be transmitted through the Internet. However, the patients’ information is related to personal privacy, therefore, patients’ information needs to be encrypted when transmited and stored. Medical image encryption is a part of it. Due to the informative fine features of medical images, a common image encryption algorithm is no longer applied. Common encryption algorithm has a single theory based on chaos image encryption algorithm, other encryption algorithms are based on information entropy. However, the images processed with these cipher text encryption algorithm are cyclical, the outline is clear and the anti-tamper capability is not strong. In view of the bit being the smallest measure unit of pixel, in order to overcome the weakness from above algorithm, and take the advantage of the chaotic system, this paper will present a chaotic medical image encryption algorithm based on bit-plane decomposition. The paper combines the image encryption and chaotic system to improve the security. This way, it can increase the security of key space and image effectively. The histogram, pixel correlation, number of pixels change rate (NPCR) and other experimental results show that the algorithm satisfies the desired effect.

Path planning of Unmanned Undersea Vehicle (UUV) is a rather complicated global optimum problem which is about seeking a superior sailing route considering the different kinds of constrains under complex combat field environment. Flower pollination algorithm (FPA) is a new optimization method motivated by flower pollination behavior. In this paper, a variant of FPA is proposed to solve the UUV path planning problem in two-dimensional (2D) and three-dimensional (3D) space. Optimization strategies of particle swarm optimization are applied to the local search process of IFPA to enhance its search ability. In the progress of iteration of this improved algorithm, a dimension by dimension based update and evaluation strategy on solutions is used. This new approach can accelerate the global convergence speed while preserving the strong robustness of standard FPA. The realization procedure for this improved flower pollination algorithm is also presented. To prove the performance of this proposed method, it is compared with nine population-based algorithms. The experiment result shows that the proposed approach is more effective and feasible in UUV path planning in 2D and 3D space.

Interval programming is one of main approaches treating imprecise and uncertain elements involved in optimization problems. In this paper, an interval linear fractional bilevel program is considered, which is characterized in that both objective coefficients and the right-hand side vector are interval numbers, and an evolutionary algorithm (EA) is proposed to solve the problem. First, the interval parameter space of the follower’s problem is taken as the search space of the proposed EA. For each individual, one can evaluate it by dealing with a simplified interval bilevel program in which only the leader’s objective involves interval parameters. In addition, the optimality conditions of linear fractional programs are applied to convert and solve the simplified problem. Finally, some computational examples were solved and the results show that the proposed algorithm is efficient and robust.