AUS Repository
Welcome to our upgraded DSpace Repository, designed to improve your user experience! Explore our improved platform and find more information on our news page.
The AUS Repository serves as the Institutional Repository of the American University of Sharjah, providing open access to research outputs from AUS students and faculty. By preserving these works for the long term and increasing their global visibility, the repository plays a key role in the dissemination of knowledge. See our About Us page for more information.
Featured Items
Recent Submissions
Item Machine Learning Based Palm Farming: Harvesting and Disease Identification(IEEE Xplore, 2024)In the culturally and economically vital date palm sector of the Arab world, precise assessment of fruit maturity, type, and disease is crucial for optimizing yield, quality, and palm health. This work pioneers a novel paradigm: machine learning (ML) frameworks for analysis of all three aspects using individual and merged datasets. Moreover, explainable AI (XAI) techniques are exploited to enhance result interpretability which has not been previously explored in this field. The purpose of this work is two-fold: 1) date fruit bunch type and ripeness classification, 2) classification of healthy and three stages of white-scale disease (WSD) infested date palm leaflets. For this purpose, we utilize deep learning (DL) models by adding additional layers and optimizing various parameters to enhance their performance for these specific tasks. Two publicly available datasets are used for both type and ripeness classification: Dataset 1 contains 8079 images, and Dataset 2 contains 9092 images of date fruit bunches. Furthermore, dataset 3 with 2161 images is used for healthy and WSD infestation stage identification. For individual datasets, the best performing model, VGG16, achieved the highest accuracy for date type classification (98%) and ripeness classification (93%), using dataset 1. The best performing classifier architecture on merged dataset, VGG16, achieved an accuracy of 97% and 94% for date fruit type and ripeness classification, respectively. The highest accuracy achieved for healthy and WSD classification was 99.7% using VGG16. These results were explained using several XAI techniques which were found to be useful in enhancing the models’ interpretability. Through this work, precision agriculture in the date palm sector stands to benefit from informed decision-making, optimized resource allocation, and the adoption of sustainable practices. This work contributes significantly to the sector's advancement, ensuring a thriving and resilient date palm industry in the region.Item Motion Images with Positioning Information and Deep Learning for Continuous Arabic Sign Language Recognition in Signer Dependent and Independent Modes(Institute of Electrical and Electronics Engineers (IEEE), 2024)While recognition of sign language alphabets and isolated words has matured in recent years, recognition of sign language sentences, or continuous signing, is still a research topic of interest in computer vision, especially in signer independent mode of recognition. Existing state-of-the-art solutions in the continuous Arabic Sign Language Recognition (ArSLR) are promising; however, when implemented in signer independent mode, the accuracies drop noticeably. In this paper, we propose a solution for recognizing continuous Arabic signing in signer dependent and independent modes through the use of motion images with positioning information. Initially, sign videos are converted into several motion images, each emphasizing a different segment of the sentence. This is achieved by a weighted sum of residual images after applying optical flow and motion compensation. Each motion image is composed of the whole sentence video, hence putting the emphasized segment into context in terms of previous and successive sign words. Thereafter, hand-crafted features are calculated from each resultant image, including numerical summaries of the horizontal, vertical and diagonal profiles. With such features, the architecture used for model generation and testing is simplified, where it consists of a single Bi-LSTM layer followed by dropout, softmax, and classification layers. This paper makes use of a recent Arabic Sign Language dataset known as ArabSign. The dataset is composed of 6 signers and 93 words arranged into 50 sentences with 30 repetitions each. Experimental results revealed that the proposed solution is suitable for signer dependent and signer independent modes of continuous sign language recognition. Using a Leave-One-Signer-Out policy, the proposed solution achieved word-recognition rates of 99.8% and 75.3%, respectively. These results noticeably surpass relevant state-of-the-art solutions.Item Characterization of PM₂̣ ₅ at a Traffic Site Using Several Integrated Analytical Techniques(Wiley, 2020)We have conducted a comprehensive, year-long, sampling campaign for particulate matter (PM) at a site near a major highway, following standard protocols. Total mass, elemental and chemical composition of the fine fractions (PM₂ ̣₅) of particulates originating from traffic are determined using several complementary techniques. These complementary techniques include gravimetric analysis, X-ray fluorescence, scanning electron microscopy and energy dispersive spectroscopy, X-ray diffraction (XRD) and black carbon multiwavelength absorption. Conducting an enrichment factor analysis and correlation coefficient calculations on elements show that Si, Ca, Al, Fe, Ti, Mn, Mg, K, Na and Cr are of crustal origin, while P, Cl and V are enriched slightly from human activities. All other measured elements (Rb, Zr, Ba, Sr, S, Ni, Cu, Zn and Pb) have high enrichment factors and relate to anthropogenic sources. Sulfates in the form of mascagnite and koktaite had the largest contribution to PM₂ ̣₅ (43% of total PM concentration). Natural pollutants such as quartz, calcite, iron oxide and aluminum oxide originating from the crust also contribute to PM₂ ̣₅ eBC and elements such as Zn, Ba, Cu, Fe and S are related to traffic emissions such as exhaust emissions and tire, brakes and road erosion. Correlation coefficients and enrichment factor calculations helped identify elements that are related to natural emissions and those related to anthropogenic sources. Being an arid region, the PM₂ ̣₅ mass concentrations were found to be within or slightly above international air quality standards.Item Source apportionment of PM₂̣ ₅ and PM₁₀ pollutants near an urban roadside site using positive matrix factorization(Elsevier, 2024-07)This paper presents results from a comprehensive study of source apportionment of particulate matter (PM) of size PM₂̣ ₅ and PM₁₀ near a busy highway in Sharjah, United Arab Emirates. Source apportionment was carried out using US Environmental Protection Agency (EPA) Positive Matrix Factorization model. Furthermore, backward trajectory analysis and Potential Source Contribution Function were used to assess air mass transport pathways and identify potential source regions, respectively. The results revealed six major sources for PM₂̣ ₅, including traffic, sea salt, fugitive dust, secondary aerosols, heavy oil combustion and mineral dust. For PM₁₀, four major sources were identified, including secondary aerosols, traffic, sea salt and mineral dust. Traffic emissions were found to be significant contributors to both PM₂̣ ₅ and PM₁₀ pollution, along with natural sources like sea salt and mineral dust. Backward trajectory analysis indicated the influence of different wind regimes on air mass transport, with contributions from regions like Arabian Gulf, Arabian Sea, Oman and Iran. The Conditional Bivariate Probability Function analysis further explained the impact of local traffic emissions and other sources on PM pollution under varying wind conditions.Item In-situ grown ternary metal hydroxides@3D oriented crumpled V₂C MXene sheets for improved electrocatalytic oxygen evolution reaction(Elsevier, 2024-08)High valence multi transition metal hydroxides are greatly enriched with OER redox active sites due to strong synergy of heteroatomic nuclei. The efficiency of these redox active sites could be efficiently improved by coupling with highly conductive substrate. The advanced three-dimensional (3D) architecture and hydrophilic terminal functionalities of MXene (MX) considerably enhance the maximum utilization rate of anchored redox active sites by triggering the direct growth of these at MX substrate. Here-in, the freeze-dried 3D network of crumpled Vanadium-Carbide (V₂C) MX sheets regulates the crystallization of in-situ grown NiFeCr multi transition metal hydroxides on MX scaffold through co-precipitation process. The XPS results suggest a synergistic chemical interaction of 3D MX scaffold with NiFeCr that modifies the electronic structure of the composite ensuring reduced charge transfer resistance. Besides, as found in FESEM morphological investigation, the well-dispersed NiFeCr multi-transition metal hydroxides are immobilized on open pores like structure of V₂C-MX facilitate thoroughly accessible active sites. As a result, the NiFeCr@3D V₂C-MX composite has shown an excellent electrocatalytic activity with an overpotential of 410 mV at a current density of 200 mA cm⁻², Tafel slope of 100 mV dec in 1M KOH. Besides, the significant interaction between metallic centers and MXene support prevent detachment or agglomeration of active centers providing maximum interaction with the electrolytic ions, quick ionic OH⁻ transportation, speedy and stable electron transfer channels thus ensure the long-term stability of NV-5MX during 53 h continuous operation of OER. Furthermore, we have utilized a more accurate value of half-cell standard reduction potential of the Hg/HgO electrode in the Nernst equation to represent all test voltages and to determine the overpotential values. In essence, this study features a facile approach for the confined growth of multi transition metal hydroxides in the presence of morphologically unique 3D crumpled V₂C MXene architectures. Consequently, the increased OER reaction kinetics and improved stability of the synthesized composites are potentially due to synergistic interplay between well dispersed active sites and the conductive substrate.
Communities in AUS Repository
Select a community to browse its collections.