Search

In this paper, we propose a dual leap motion controllers (LMC)-based Arabic sign language recognition system. More specifically, we introduce the concept of using both front and side LMCs to cater for the challenges of finger occlusions and missing data. For feature extraction, an optimum set of geometric features is selected from both controllers, while for classification, we used both a Bayesian approach with a Gaussian mixture model (GMM) and a simple linear discriminant analysis (LDA) approach. To combine the information from the two LMCs, we introduce an evidence-based fusion approach; namely, the Dempster–Shafer (DS) theory of evidence. Data were collected from two native adult signers, for 100 isolated Arabic dynamic signs. Ten observations were collected for each of the signs. The proposed framework uses an intelligent strategy to handle the case of missing data from one or both controllers. A recognition accuracy of about 92% was achieved. The proposed system outperforms state-of-the-art glove-based systems and single-sensor-based techniques.

In critical applications, such as medical diagnosis, security related systems, and so on, the cost or risk of action taking based on incorrect classification can be very high. Hence, combining expert opinions before taking decision can substantially increase the reliability of such systems. Such pattern recognition systems base their final decision on evidence collected from different classifiers. Such evidence can be of data type, feature type, or classifier type. Common problems in pattern recognition, such as curse of dimensionality, and small sample data size, among others, have also prompted researchers into seeking new approaches for combining evidences. This paper presents a criteria-based framework for multi-classifiers combination techniques and their areas of applications. The criteria discussed here include levels of combination, types of thresholding, adaptiveness of the combination, and ensemble-based approaches. The strengths and weaknesses of each of these categories are discussed in details. Following this analysis, we provide our perspective on the outlook of this area of research and open problems. The lack of a well-formulated theoretical framework for analyzing the performance of combination techniques is shown to provide a fertile ground for further research. In addition to summarizing the existing work, this paper also updates and complements the latest developments in this area of research.

Accurate wind energy assessments require wind speed (WS) at the hub height. The cost of WS measurements grows enormously with height. This paper utilizes deep neural network (DNN) algorithm for the extrapolation of the WS to higher heights based on measured values at lower heights. LiDAR measurements at lower heights are used for training the system and at higher heights for performance analysis. These measurements are made at 10, $20,\ldots $ , and 120 m heights. First, the measured WS values at 10–40 m were used to extrapolate values up to 120 m. In the second scenario, the WS at 10–50 m were used to extrapolate values up to 120 m. This continued until the last scenario, in which the WS at 10–100 m were used to estimate values at 110 and 120 m. A relationship between heights of measurements and the accuracy of the WS estimation at hub height is presented. The WS extrapolated using the present approach is compared with the measured values and with local wind shear exponent (LWSE)-based extrapolated WS. Furthermore, to analyze the performance of the DNN relative to other machine learning methods, we compared its performance with that of classical feedforward artificial neural networks trained using a genetic algorithm to find the initial weights and the Levemberg–Marquardt (LM) method (GANN) for training. The mean absolute percent error between measured and extrapolated WS at height 120 m based on measurements between 10–50 m using DNN, GANN, and LWSE are 9.65%, 12.77%, and 9.79%, respectively.

Monitoring students’ class attendance in any educational institution is an important process as it is directly linked to academic performance. Collecting the student attendance manually results in ...

Several million people from around the world visit Saudi Arabia each year to perform Hajj (pilgrimage). One of the largest socio-religious gatherings in the world, the Hajj poses unique challenges in terms of safety, security, crowd control, medical emergencies, transportation, and a variety of other logistical considerations. To add further complexity, these visitors come from a myriad of countries around the globe, representing a variety of diverse cultures and speaking more than 120 distinct languages. Moreover, the majority of these pilgrims are of older age, and prone to falling ill and losing money and identification documents. Several die naturally during these few days of pilgrimage every year. Billions of dollars are being spent to improve safety, security, and infrastructure, as well as to provide better services in hopes of making the event a memorable one for pilgrims. However, several challenges remain, many of which can be overcome by implementing the latest technologies to make this Holy trip easier on pilgrims and authorities alike. In this article, the benefits of employing the novel technology of Near Field Communication (NFC) to manage large socio-religious events such as Hajj are described. NFC-enabled mobile devices can expedite identity screening and store medical records of pilgrims for timely aid in the event of an emergency, be used as a transportation pass between Holy sites, and function as official documents for pilgrims so these won?t get lost. Several of these services can be provided even if a pilgrim does not have an NFC-enabled mobile phone by providing pre-programmed NFC tags at the point of arrival.

In this paper, we propose a Dual-Leap Motion Controllers (DLMC) based Arabic Sign Language recognition system. More particularly, we propose to use both front and side controllers to cater for the challenges of finger occlusions and missing data. For feature extraction, we select an optimum set of geometric features extracted from both controllers, while for classification, we used both a Bayesian approach with a Gaussian Mixture Model (GMM) and a simple Linear Discriminant Analysis (LDA) approach. Though this paper focused only on the GMM approach. Data was collected from a native adult signer, for 100 isolated Arabic words. Ten observations were collected for each of the signs. The proposed framework uses an intelligent strategy to handle the case of missing data from one or both controllers. A recognition accuracy of 94.63% was achieved, with the proposed system. The proposed system outperforms glove-based systems and a single-LMC based techniques.

In this paper, three methods are used for the prediction of wind speed, 12 h ahead, based on 72 h previous wind speed values at three locations viz. Rawdat Bin Habbas (inland north), Juaymah (east coast), and Dhulom (inland western region) in Saudi Arabia. These methods are Particle Swarm Optimization (PSO), Abductory Induction Mechanism (AIM), and the Persistence (PER) model. The available data at each site was divided into three consecutive groups. The first 50% was used for training, the second 25% for validation, and the remaining 25% for testing. The validation data set was used to select the network architecture and other user defined parameters. The testing data was used only to assess the performance of the networks on future unseen data that has not been used for training or model selection. For each of the three methods, each of 12 networks was trained to produce the wind speed at one of the next 12 h. Relatively, Close agreements were found between the predicted and measured hourly mean wind speed for all three locations with coefficient of correlation R 2 values between 81.7% and 98.0% for PSO, between 79.8% and 98.5% for AIM and between 59.5% and 88.4% for persistence model. Both PSO and AIM methods underestimated WS values during most hours with an average value of 0.036 m/s and 0.02 m/s, respectively. However, persistence model overestimated the WS by an average value of 0.51 m/s. It is shown that the two developed models outperformed the persistence model on predicting wind speed 12 h ahead of time with slight advantage to the PSO method.

Several studies have been carried on sign language recognition systems, however, practically deployable system for real-time use is still a challenge. Traditionally, sign language recognition systems have either used sensor gloves or digital cameras to acquire and process hand gestures. Both approaches exhibit some disadvantages for real time deployment that hindered it large scale adoption. With the growth witnessed in gaming systems, two new instruments have been introduced namely, the Microsoft Kinect (MK) and the leap motion controller. The MK system has been developed to interact with video games by tracking full body movements and gestures. To overcome some of the disadvantages of the classical methods, we propose here to develop an Arabic sign language recognition system based on MK system. The developed system was tested with 20 signs from the Arabic sign language dictionary. Therefore, in this paper, we present our experiment carried out using the MK setup on 20 Arabic sign language words. Video samples of both true color images and depth images were collected from native deaf signer. Linear Discriminant analysis was used for feature dimension reduction and sign classification. Furthermore, fusion from RGB and depth sensor was carried at feature and decision level giving an overall best accuracy of 99.8%.

Sign language maps letters, words, and expressions of a certain language to a set of hand gestures enabling an individual to communicate by using hands and gestures rather than by speaking. Systems capable of recognizing sign-language symbols can be used for communication with the hearing-impaired. This paper represents the first attempt to recognize two-handed signs from the Unified Arabic Sign Language Dictionary using the CyberGlove and support vector machines (SVMs). 20 samples from each of 100 two-handed signs were collected from two adult signers. Because the signs are of different lengths, time division is used to standardize sign length. The duration of every sign is divided into a specific number of segments, and the mean and standard deviation of each segment are used to represent the signal in the segment. After pre-processing, principal component analysis is used for feature extraction. For recognition, a SVM is trained on 15 samples from each sign. The performance is obtained by testing the trained SVM on the remaining five samples from each sign. A recognition rate of 99.6% on the testing data is obtained.

The quantity of solar radiation received by the earth’s surface is very important to numerous renewable energy applications. However, direct measurement of solar data is not widely available, especially in developing countries. This paper uses Particle Swarm Optimization (PSO) to train an artificial neural network (PSO–ANN) using data from available measurement stations to estimate monthly mean daily Global Solar Radiation (GSR) at locations where no measurement stations are available. The inputs to the networks are: month of the year, latitude, longitude, altitude, and sunshine duration, and the output is the monthly mean daily GSR at the specified location. Using training data from 31 stations and testing data from 10 locations, the PSO–ANN outperforms a neural network trained using the standard backpropagation (BP) algorithm (BP–ANN) with an average Mean Absolute Percentage Error (MAPE) of 8.85% for the PSO–ANN and 12.61% for the BP–ANN. The performance is improved significantly, when we use the leave-one-out method, where data from 40 locations is used for training and data from the 41st station is used for assessing the performance. In this case the average of MAPE on data from the 10 testing stations is about 7%. We used the same method to assess the performance of the PSO–ANN on testing data from each of the 41 stations with an overall average MAPE of about 10.3%. Comparison with BP–ANN and an empirical model showed the superiority of the PSO–ANN.

The present study utilizes the measured global solar radiation, ambient temperature, and relative humidity data from Al-Hasa, Al-Jouf, and Sharurah radiation data collection stations between 1998 and 2002 to estimate the fractions of diffuse solar radiation and direct normal radiation. For estimation purposes, a radial basis function neural network has been utilized. Specifically, the reported work developed a model with a four input parameter model, i.e., day of the year, global solar radiation, ambient temperature, and relative humidity for the estimation of diffuse solar radiation and direct normal radiation fractions of solar radiation from global solar radiation. The models so developed were trained with the measured data during the period 1998 to 2001, while the data for the year 2002 was used for testing the model results.

SUMMARY This study utilizes Abductory Induction Mechanism to estimate the mean monthly wind speed at some locations in Saudi Arabia based on wind data at other available recording stations in addition to some historical wind speed data at the target site. Wind speed data from 20 meteorological stations over a period of 16 years between 1990 and 2005 was used to accomplish the set objective. To validate the model, data from 19 stations was used to estimate the wind speed at the 20th location. Evaluation was performed for every one of the 20 available locations. Results show good agreement between estimated and measured monthly mean wind speed values. Copyright © 2010 John Wiley & Sons, Ltd.

Sign language is important for facilitating communication between hearing impaired and the rest of society. However, most vocal people do not understand sign language, hence, the need to develop system capable of translating sign language. Two approaches have traditionally been used in the literature: image-based and glove-based systems. Glove-based systems require the user to wear electronic gloves while performing the signs. The glove includes a number of sensors detecting different hand and finger articulations. Image-based systems use camera(s) to acquire a sequence of images of the hand. Each of the two approaches has its own disadvantages. The glove-based method is not natural as the user must wear a cumbersome instrument while the camera-based system requires specific background and environmental conditions to achieve high accuracy. In this paper, we propose a new approach for Arabic Sign Language Recognition (ArSLR) which involves the use of two Leap Motion Controllers (LMC) to prevent the case of one finger being occluded by another finger or hand. This device detects and tracks the hand and fingers to provide position and motion information. We propose to use the two LMCs as a backbone of the ArSLR system. In addition to data acquisition, the system includes a preprocessing stage, a feature extraction stage, and a classification stage. Fusion of evidences from the two LMCs at the feature extraction and classification stage was also investigated using Dempster-Shafer theory of evidence. Features fusion from the two LMCs gives 97.7% classification accuracy with Linear Discriminant Analysis (LDA) classifier and 97.1% with classifier level fusion. This gives better recognition over the use of a single LMC.

IMPORTANCE: Lipoprotein-associated phospholipase A2 (Lp-PLA2) has been hypothesized to be involved in atherogenesis through pathways related to inflammation. Darapladib is an oral, selective inhibitor of the Lp-PLA2 enzyme. OBJECTIVE: To evaluate the efficacy and safety of darapladib in patients after an acute coronary syndrome (ACS) event. DESIGN, SETTING, AND PARTICIPANTS: SOLID-TIMI 52 was a multinational, double-blind, placebo-controlled trial that randomized 13,026 participants within 30 days of hospitalization with an ACS (non-ST-elevation or ST-elevation myocardial infarction [MI]) at 868 sites in 36 countries. INTERVENTIONS: Patients were randomized to either once-daily darapladib (160 mg) or placebo on a background of guideline-recommended therapy. Patients were followed up for a median of 2.5 years between December 7, 2009, and December 6, 2013. MAIN OUTCOMES AND MEASURES: The primary end point (major coronary events) was the composite of coronary heart disease (CHD) death, MI, or urgent coronary revascularization for myocardial ischemia. Kaplan-Meier event rates are reported at 3 years. RESULTS: During a median duration of 2.5 years, the primary end point occurred in 903 patients in the darapladib group and 910 in the placebo group (16.3% vs 15.6% at 3 years; hazard ratio [HR], 1.00 [95% CI, 0.91-1.09]; P = .93). The composite of cardiovascular death, MI, or stroke occurred in 824 in the darapladib group and 838 in the placebo group (15.0% vs 15.0% at 3 years; HR, 0.99 [95% CI, 0.90-1.09]; P = .78). There were no differences between the treatment groups for additional secondary end points, for individual components of the primary end point, or in all-cause mortality (371 events in the darapladib group and 395 in the placebo group [7.3% vs 7.1% at 3 years; HR, 0.94 [95% CI, 0.82-1.08]; P = .40). Patients were more likely to report an odor-related concern in the darapladib group vs the placebo group (11.5% vs 2.5%) and also more likely to report diarrhea (10.6% vs 5.6%). CONCLUSIONS AND RELEVANCE: In patients who experienced an ACS event, direct inhibition of Lp-PLA2 with darapladib added to optimal medical therapy and initiated within 30 days of hospitalization did not reduce the risk of major coronary events. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT01000727.

Most sign language recognition systems that use gloves and hand trackers combine the data from both devices at the sensor level. In this paper we propose a new approach by combining information acquired from the gloves and the hand tracking systems at the decision level using the Dempster-Shafer theory of evidence. The results using the Dempster-Shafer on the recognition of 100 two-handed signs show enhanced performance compared to the individual systems and to classification based on combined features. A recognition accuracy of 84.7%, and 91.3% are achieved when attempting to recognize the signs from the hand tracker only, and the glove data, respectively. When the sensor data from the gloves and hand tracking systems are combined, a recognition accuracy of 96.2% was achieved while a recognition accuracy of 98.1% was achieved when the fusion is performed at the decision level using Dempster-Shafer theory of evidence.

RFID (radio frequency identification) and license plate recognition (LPR) technologies have proven to be very effective in providing innovative solutions to numerous real world applications. In this paper, we propose an RFID-based system supported by license plate recognition (LPR) technology for monitoring and controlling entry of vehicles to the Holy region of Makkah (Saudi Arabia) during Hajj (Pilgrimage) season. Authorized vehicles are provided with passive RFID tags that are programmed to specify their schedule of entry. This is useful in allowing entry of service vehicles during low traffic periods of pilgrims' vehicles. The proposed LPR system is based on a two-stage license plate-character recognition algorithm. The license plate region is detected using morphological filtering while the character recognition stage uses a neural network approach. The extensive tests we conducted showed that the system is robust to different traffic conditions and is able to limit unauthorized vehicles from entry during Hajj period. A vehicle recognition accuracy of around 95% was achieved under normal conditions.

In high-risk hypertensive subjects (HTs) with incidental unilateral renal artery stenosis (RAS), the effectiveness of percutaneous revascularization with stent (PR-STENT) on blood pressure (BP) and glomerular filtration rate (GFR) is not established.Eighteen HTs aged 65.7 ± 9.2 years with angiographically diagnosed unilateral RAS (≥ 60%) were randomized to receive PR-STENT (N=9) or to NO-STENT (N=9). BP (mercury sphygmomanometer) and GFR (99mTc-DTPA clearances during renal scintigraphy) were evaluated yearly for three years. Echo-Doppler of renal arteries was performed to verify the anatomic patency and flow velocities of the reperfused artery. Analysis of variance compared BP and GFR values changes from baseline to the follow-up; differences for continuous variables were evaluated between groups with the Tukey's post hoc test after adjustment for age, change of BP between baseline and at the follow-up, GFR and body mass index (BMI).Baseline systolic BP and GFR values were not different between groups. The significantly greater GFR increase observed in PR-STENT than in NO-STENT at univariate analysis at the end of follow-up (62.5 ± 19.2 vs. 42.24 ± 17.6, P0.02) disappeared after adjustment for confounding factors. However, systolic BP remained significantly lower in PR-STENT than in NO-STENT (140.1 ± 4.6 vs. 170.0 ± 8.3, P0.0001) also after adjustment for age, GFR and BMI.PR-STENT reduces systolic BP without improving GFR. Due to the strong association between high BP and renal damage, this study raises the question on whether PR-STENT should be performed in all HTs with unilateral and incidental RAS.

This paper describes a system developed for the tracking and identification of pilgrims in the Holy areas, in Makkah-Saudi Arabia, during Hajj (Pilgrimage). The area is already covered by a sophisticated 3.5G network by several service providers. Upon request or periodically, the mobile phone sends its UID, latitude, longitude, and time stamp. A server maps the latitude and longitude information on a Google map or any geographical information system. If the Internet connection is lost the mobile phone stores the location information in its memory until the Internet connection is restored, then it sends all stored location information and clears this information from memory. The developed system can be used to track a specific pilgrim. Alternatively any pilgrim can request emergency help using the same system. The developed system works in coordination with an RFID identification system that was proposed earlier. The developed system was tested successfully during the recent pilgrimage season.

In this paper we propose machine learning algorithms to estimate the sunshine duration at any region in Saudi Arabia using easily obtained parameters that include: maximum possible day length (So), extraterrestrial solar radiation at that particular location (Ho), latitude, longitude, altitude, and month number. Two neural network algorithms, namely, Particle Swarm Optimization (PSO) algorithm and Support Vector Machine (SVM), are used to estimate the sunshine duration (S). Due to the limited number of measurement stations, we use the leave-one-out strategy, where data from 40 measurement stations are used to build the model and data from the remaining 41st station are utilized to assess accuracy of the modeled system. This process is repeated for each of the 41 stations. Minimum mean absolute percent errors (MAPE) of 2.3% and 2.7% were obtained at Al-Madina station using PSO and SVM methods while the respective maximum values of 22.9% and 16.7% were observed at Al-Numas station. In general, higher MAPE values were found during summer time (June-August) and lower in winter time (January-February and November-December). Results indicate the viability of the proposed methods for estimating sunshine duration in Saudi Arabia, and the system is expected to perform similarly in other countries with similar environment.

Sign language is important for facilitating communication between hearing impaired and the rest of society. Two approaches have traditionally been used in the literature: image-based and sensor-based systems. Sensor-based systems require the user to wear electronic gloves while performing the signs. The glove includes a number of sensors detecting different hand and finger articulations. Image-based systems use camera(s) to acquire a sequence of images of the hand. Each of the two approaches has its own disadvantages. The sensor-based method is not natural as the user must wear a cumbersome instrument while the imagebased system requires specific background and environmental conditions to achieve high accuracy. In this paper, we propose a new approach for Arabic Sign Language Recognition (ArSLR) which involves the use of the recently introduced Leap Motion Controller (LMC). This device detects and tracks the hand and fingers to provide position and motion information. We propose to use the LMC as a backbone of the ArSLR system. In addition to data acquisition, the system includes a preprocessing stage, a feature extraction stage, and a classification stage. We compare the performance of Multilayer Perceptron (MLP) neural networks with the Nave Bayes classifier. Using the proposed system on the Arabic sign alphabets gives 98% classification accuracy with the Nave Bayes classifier and more than 99% using the MLP.