Home Enhancing communication: Deep learning for Arabic sign language translation
Article Open Access

Enhancing communication: Deep learning for Arabic sign language translation

  • Mosab A. Hassan EMAIL logo , Alaa H. Ali and Atheer A. Sabri
Published/Copyright: July 2, 2024
Download Article (PDF)
Become an author with De Gruyter Brill
Submit Manuscript Author Information
Open Engineering
From the journal Open Engineering Volume 14 Issue 1

Abstract

This study explores the field of sign language recognition through machine learning, focusing on the development and comparative evaluation of various algorithms designed to interpret sign language. With the prevalence of hearing impairment affecting millions globally, efficient sign language recognition systems are increasingly critical for enhancing communication for the deaf and hard-of-hearing community. We review several studies, showcasing algorithms with accuracies ranging from 63.5 to 99.6%. Building on these works, we introduce a novel algorithm that has been rigorously tested and has demonstrated a perfect accuracy of 99.7%. Our proposed algorithm utilizes a sophisticated convolutional neural network architecture that outperforms existing models. This work details the methodology of the proposed system, which includes preprocessing, feature extraction, and a multi-layered CNN approach. The remarkable performance of our algorithm sets a new benchmark in the field and suggests significant potential for real-world application in assistive technologies. We conclude by discussing the impact of these findings and propose directions for future research to further improve the accessibility and effectiveness of sign language recognition systems.

Keywords: sign language recognition; principal component analysis; linear discriminant analysis; deep learning; convolutional neural network

1 Introduction

Deaf individuals often rely on sign language for everyday communication. This unique form of communication, though prevalent within the deaf community, remains relatively rare outside of it. This leads to significant communication barriers between deaf and hearing people. For instance, hearing parents with deaf children may face difficulties due to the language barrier. This challenge extends to raising, nurturing, and imparting Islamic traditions to deaf children [1]. Various Arabic sign languages (ArSLs), including Egyptian, Jordanian, Tunisian, and Gulf sign languages, utilize a shared alphabetic system. However, deaf individuals face obstacles due to a lack of accessible information and difficulties in communication, particularly in performing religious rituals. This highlights the need for machine translation solutions to bridge these gaps, allowing deaf individuals to access education and scientific knowledge in their native sign language [2,3]. Advances in pattern recognition and human-computer interaction, especially in the fields of computer vision and machine learning, are crucial. These technologies are key in recognizing hand gestures used by the deaf for Qur’anic alphabet letters.

Hearing loss is a significant global health concern, as highlighted by the World Health Organization. It affects approximately 5% of the world’s population, translating to over 460 million individuals, including 34 million children. The prevalence of hearing loss is expected to rise, with projections suggesting that nearly 900 million people could be affected by 2050. Additionally, there is a growing concern for 1.1 billion children who are at risk of hearing loss due to loud noise exposure and other factors. The economic impact is substantial, with hearing loss costing the global economy an estimated 750 billion dollars [1]. Hearing impairment is classified into various degree. Those with severe to profound hearing loss often face significant challenges in paying attention to and understanding spoken language, leading to communication barriers. These barriers can have profound implications on mental health, potentially leading to feelings of isolation, loneliness, and unhappiness in the deaf community.

To bridge the communication gap, the deaf community relies on sign language, a visual-gestural language using hand gestures, facial expressions, and body movements. However, this form of communication is not widely understood by the hearing population, further exacerbating the communication challenges between deaf and hearing individuals.

The diversity in sign languages mirrors that of spoken languages, with approximately 200 distinct sign languages globally. Each sign language has its own unique structure and lexicon, just as spoken languages vary from one region to another. This diversity not only reflects the rich cultural and linguistic tapestry of the deaf community but also underscores the complexity of facilitating effective communication across different sign languages.

Sign language serves as a vital communication tool for the deaf community. It employs a range of bodily actions, including gestures or signs, to convey messages. This method of communication is distinct from spoken languages and utilizes various physical expressions such as head nods, shoulder shrugs, hand movements, and facial expressions to relay messages. The proposed work aims to facilitate interaction within the deaf community and between deaf and hearing individuals. In sign language, each gesture represents a letter, word, or emotion, forming phrases through a combination of signs, much like words form sentences in spoken languages. This has led to the development of a complete natural language with its own sentence structure and grammar.

Deep learning (DL), a subset of machine learning algorithms, is instrumental in representing complex structures through multiple nonlinear transformations. The foundation of DL lies in neural networks, which have spurred significant advancements in fields like image and sound processing, including face and voice recognition, automated language processing, computer vision, text classification, medical diagnosis, and genomics.

DL algorithms employ computational methods to extract representations of data across multiple layers, discovering patterns in large datasets. This is achieved using backpropagation, which adjusts the internal parameters of a system for each level of representation. Deep convolutional networks have shown remarkable progress in processing videos, images, speech, and audio, while recurrent networks excel in handling sequential data like voice and text.

Neural network architecture plays a crucial role in DL. The term “deep” in DL refers to the number of layers in a network; more layers imply greater complexity and capability of the system. DL is notable for its accuracy, often surpassing human capabilities, thanks to modern tools and methods.

The ultimate goal is to develop technology capable of recognizing sign language and translating the most common gestures of deaf individuals into written data. The aim of this technological advancement is to bridge the communication gaps and facilitate better understanding and interaction between the deaf and hearing communities.

In a study highlighted by Tharwat et al. [1], researchers developed a machine learning system for recognizing the ArSL alphabet. This system was tested using 2,800 images, representing 28 alphabets, with each alphabet class represented by 10 participants. For each letter, 100 images were used, totaling 2,800 images. The system employed a feature extraction method based on hand shape, where each image was described by a 15-value vector indicating key point locations. Another approach by Sidig et al. [4] found that the Hartley transform, in combination with a SVM classifier, detected ArSLR with an impressive 98.8% accuracy. Alzhohairi et al. [5] explored an image-based method to recognize Arabic alphabet movements, achieving a 63.5% success rate. Kamruzzaman [6], in 2020, introduced a vision-based method for identifying Arabic hand signs and converting them to Arabic speech. This method, using a Convolutional Neural Network (CNN), reported a 90% recognition rate. Similarly, Elbadawy et al. [7] proposed a CNN framework to recognize 25 ArSL signs, achieving training and testing accuracies of 85 and 98%, respectively. Mohamed [8] discussed a system using depth-measuring cameras and computer vision techniques for capturing and segmenting images of facial expressions and hand gestures, with a 90% recognition rate.

Researchers have explored various CNN architectures for sign language recognition, as detailed in several studies. A previous study [9] analyzed the impact of dataset size on CNN model accuracy using a collection of 54,049 sign images. In the study by Latif et al. [10], they found that increasing the dataset size significantly enhances model accuracy, noting an improvement from 80.3 to 93.9% with larger datasets. Further accuracy gains were observed when dataset sizes varied between 33,406 and 50,000 samples, elevating accuracy from 94.1 to 95.9%.

In another investigation [11], a novel CNN architecture, ArSL-CNN, was developed for Arabic sign language recognition using the ArSL2018 dataset. The initial training and testing accuracies of the ArSL-CNN were 98.80 and 96.59%, respectively. The study also examined the effect of data resampling techniques, such as the synthetic minority oversampling method (SMOTE), to address data imbalances, ultimately improving testing accuracy to 97.29%.

A different approach was taken in research [12], where transfer learning and deep CNN fine-tuning were applied to the same ArSL2018 dataset. This was aimed at enhancing the recognition accuracy of 32 hand motions. To address class size disparities, random under sampling was employed, reducing the total image count to 25,600.

Further, a deep transfer learning-based method for ArSL was proposed in the study [13]. This method utilized data augmentation and fine-tuning techniques within the transfer learning framework to minimize overfitting, achieving a notable accuracy of 99.52% with the ResNet101 network.

Another research [14] introduced an innovative system for translating Ethiopian sign language (ETHSL) into Amharic alphabets. This system, which employed deep CNNs and computer vision techniques, consisted of preprocessing, feature extraction, and recognition stages.

Finally, a study [15] explored the development of an autonomous translator for Amharic sign language using digital image processing and machine learning techniques. This system extracted 34 features from hand motions, including shapes, colors, and movements, and utilized ANN and multiclass SVM classifiers. The summarization of the related work in sign language recognition research is presented in Table 1.

Table 1

Summary of recent research in sign language recognition: Comparative analysis of methodologies and accuracies

Study reference Focus area Methodology Key features Accuracy%
[1] Arabic sign language alphabet recognition Machine learning with KNN and MLP Hand shape-based feature extraction with 15 values vector 97.548
[4] ArSL recognition (ArSLR) Fourier, Hartley, and Log-Gabor transforms with SVM Hartley transform for ArSLR detection 98.8
[5] Arabic alphabet movement recognition Image-based method — 63.5
[6] Arabic hand sign recognition and conversion to speech Vision-based approach with CNN — 90
[7] Recognition of 25 ArSL signs CNN framework — 85–98% (training-testing)
[8] Facial expressions and hand gestures capture Depth-measuring cameras and computer vision — 90
[10] Various CNN architectures for sign language CNN with large dataset Dataset size impact study 80.3–97.6 (increasing with dataset size)
[11] ArSL recognition with ArSL-CNN Novel ArSL-CNN architecture Use of SMOTE for imbalanced data 96.59–97.29 (initial-post SMOTE)
[12] Hand motion recognition from ArSL-CNN Transfer learning and deep CNN fine-tuning Dataset size reduction for class size disparity 99.4–99.6 (VGG-16 and ResNet-152)
[13] ArSL identification Deep transfer learning with ResNet101 Fine-tuning and data augmentation 99.52
[14] ETHSL to Amharic alphabets translation Deep CNN and computer vision Comprehensive system involving preprocessing, feature extraction, and recognition 98.3 (testing)
[15] Autonomous Amharic sign language translator Digital image processing and machine learning ANN and multiclass SVM classification 80.82–98.06 (ANN-SVM)

2 Proposed methodology

The methodology depicted in Figure 1 outlines a structured process for recognizing ArSL from images, which we have delved into more thoroughly in Sections 2.1–2.4. Initially, the process begins with preprocessing the input data from the ArSL2018 dataset. During this stage, images are first converted to grayscale to reduce complexity. Next noise is reduced by Gaussian blur and this is followed by applying a histogram equalization to enhance the contrast of the images, to ensure that the data are uniformly distributed across all intensities. The final preprocessing step involves resizing the images to a standard size, facilitating consistent input dimensions for feature extraction. In the feature extraction phase, Principal component analysis (PCA) and Linear discriminant analysis (LDA) are used. PCA reduces the dimensionality of the data by identifying the principal components that capture the most variance within the data, which simplifies the complexity while retaining significant information. LDA, on the other hand, focuses on maximizing the separability between different sign language classes to improve the classifier’s ability to distinguish between them. After preprocessing and feature extraction, the processed data are fed into the proposed CNN. The CNN architecture is designed to further analyze and learn from the data, extracting higher-level features through its multiple layers.

Figure 1 The proposed scheme.
Figure 1

The proposed scheme.

The final step in the methodology is classification, where the CNN outputs are used to categorize the images into their respective sign language classes. This step is crucial as it translates the extracted features into meaningful predictions that correspond to specific signs in the ArSL alphabet. Each of these steps contributes to the overall goal of accurately translating visual sign language data into a format that can be understood and utilized, with the aim of improving communication for the deaf community. Each of these steps are explained in more detail, as illustrated in Figure 1.

2.1 ArSL2018 dataset

The ArSL2018 dataset represents a novel and expansive collection of ArSL imagery, introduced by Prince Mohammad bin Fahd University in Al Khobar, Saudi Arabia. This dataset has been made accessible to the research community, particularly those working in Machine learning and DL, to foster advancements in assistive technology for the benefit of individuals who are deaf or hard of hearing. Comparable datasets are referenced in the studies by Latif et al. [16] and Athitsos et al. [17]. According to the creators’ understanding, the ArSL2018 stands out as the first extensive dataset dedicated to ArSL. Comprising 54,049 grayscale images, each with a resolution of 64 × 64 pixels, the ArSL2018 dataset includes a variety of images that account for different lighting conditions and backgrounds, enriching the dataset’s diversity. Figure 2 presents a subset of images from the dataset, showcasing ArSL signs and alphabets. The dataset has been meticulously curated, with images collected, labelled, and compiled, and is now available for researchers. This resource is anticipated to not only enhance the accuracy of the sign language classification and recognition algorithms but also to serve as a foundational tool for developing prototypes aimed at improving communication within the deaf community [18].

Figure 2 ArSL representation for Arabic alphabets.
Figure 2

ArSL representation for Arabic alphabets.

2.2 Preprocessing phase

During the preprocessing phase of image dataset handling, several critical steps are taken to improve the quality of images, which are essential for both training and testing models, as well as for classifying new images.

The initial operation involves converting color [19] images to grayscale, which is a vital step as it reduces the data space and simplifies subsequent processes. Color images, composed of red, green, and blue components [20] are transformed using Equation (1), Figure 3 shows the process of converting images to grayscale.

(1) Grayscale = 0.30 R + 0.59 G + 0.11 B .

Figure 3 Grayscale images.
Figure 3

Grayscale images.

The result of this conversion is an image in shades of gray, eliminating the need to process three different color channels.

Following grayscale conversion, a Gaussian blur [21] is applied to mitigate noise and blurring, which can negatively affect the model’s generalization performance. This technique uses a Gaussian function in Equation (2), Figure 4 shows the process of Gaussian blur in images

(2) G ( q ) = 1 2 π σ e − q 2 σ 2 ,

where ( σ ) is the standard deviation of the Gaussian distribution, smoothing the image by averaging the pixels based on their spatial proximity.

Figure 4 Gaussian blur effect on images.
Figure 4

Gaussian blur effect on images.

Histogram equalization [22] is then utilized to enhance the image contrast, redistributing the pixel intensity distribution to achieve a more uniform histogram. This process effectively addresses issues with lighting and background variations in images [23]. Equation (3) represents the graph equalization of the equation and Figure 5 represents the images after the graph equalization process.

(3) h [ i ] = ∑ X = 1 N ∑ Y = 1 N 0 if f [ x . y ] = i 1 otherwise .

Figure 5 Histogram equalization effect on images.
Figure 5

Histogram equalization effect on images.

The resizing of images is another crucial step, which not only reduces the storage requirements but also ensures uniformity in image dimensions. This is achieved through a bilinear interpolation method, which considers both horizontal and vertical pixel values to adjust the image to the desired resolution [24]. Equation (4) represents the resizing equation and Figure 6 represents resizing images.

(4) y = y ₒ 1 + x − x ₒ x 1 − x ₒ + y 1 1 − x 1 − x x 1 − x ₒ .

Figure 6 Resizing effect on the image.
Figure 6

Resizing effect on the image.

2.3 Feature extraction phase

In the feature extraction section of our study on ArSL, we employ two critical algorithms to refine the data and enhance classification performance: PCA [25,26] and LDA [27].

PCA is a statistical method that reduces the dimensionality of a dataset consisting of related variables while ensuring that these new variables are independent of each other. The essence of PCA is to capture as much information as possible with fewer features. This is achieved through a transformation that identifies the patterns in the data. The key aspects of PCA involve calculating the mean of the training images, which is given by Equations (5) and (6).

(5) Avarage = 1 M ∑ n = 1 μ training images ( n ) ,

(6) Cov = ∑ n = 1 μ sub ( n ) sub T ( n ) ,

where M is the total images training set, μ represents the average mean, and Sub represents the image that is subtracted from the average μ.

Following this, the covariance is determined, representing the variance between each data point in the training set.

LDA is a well-regarded statistical technique in pattern recognition and machine learning applications due to its ability to reduce the dimensions of images while maintaining the characteristics necessary for accurate recognition. LDA seeks a lower-dimensional space where projections of feature vectors are well separated for each class. This method calculates the mean vectors for each category, overall mean value, between-class and within-class scatter matrices, followed by the extraction of linear discriminants. The procedure is outlined by a series of Equations (7)–(12) which define these concepts.

(7) μ j = 1 n j ∑ x i ϵ w i x i ,

(8) μ = 1 N ∑ i = 1 N x i = ∑ i = 1 c n i N μ ,

(9) SB = ∑ i = 1 N n i ( μ i − μ ) ( μ i − μ ) ᵀ ,

(10) S W = ∑ j = 1 c ∑ i = 1 n i ( x ij − μ i ) ( x ij − μ j ) ᵀ ,

(11) W = S − 1 w SB ,

(12) Y = XV ₖ ,

where N is the total number of samples.

In the context of linear discriminant analysis, μ i denotes the mean vector of the ith class derived from the dataset. This mean vector represents a central point that characterizes the ith class in terms of its features. The term ‘projection’ here refers to the transformation of this class mean into a new coordinate system defined by the linear discriminants. This projection helps to maximize the separation between classes while minimizing the variance within each class, facilitating more effective classification.

SW i the within-class variance of the ith class, represents the difference between the mean.

We apply these methodologies in our work to effectively reduce the complexity of the images in the ArSL, aiming to discern distinct patterns that aid in recognition. Through PCA, we diminish the redundant data, and with LDA, we ensure that the resulting features are optimal for distinguishing between different signs. This dual approach enhances the model’s ability to classify the sign language images accurately, thus contributing to the development of more robust recognition systems for assisting the deaf community [28].

2.4 The proposed CNN

The architecture of CNN represents a significant advance in the field of DL [29] particularly in tasks related to image processing. CNNs operate by extracting features directly from pixel data, a process that allows for high-level representation of images. The core operation within a CNN is known as convolution, a specialized kind of linear operation that processes data through a series of matrix manipulations, Figure 7.

Figure 7 The proposed CNN.
Figure 7

The proposed CNN.

CNNs have demonstrated exceptional performance in areas such as image classification, object detection, and even the recognition of complex behaviors. This success is largely due to the availability of large-scale datasets that contain millions of labelled examples, providing the extensive data necessary for training these DL models.

In the CNN structure proposed for our work, the model processes input images initially sized at 64 × 64 pixels, which, after preprocessing and feature extraction, are resized to 20 × 20 pixels. These images are then converted into one-dimensional vectors to facilitate the detection of hand gestures. The convolutional layers within the network are responsible for extracting pertinent features from the input image by sliding a filter over the image and applying a convolution operation at each position.

Following the convolutional layers, the network employs Max Pooling layers to reduce the dimensionality of the feature maps while maintaining the depth, which corresponds to the number of channels. To address potential issues where neurons could become inactive – a problem known as “dying ReLU” – Leaky ReLU activation functions are used, which allow a small gradient when the unit is not active.

To capture more complex patterns, especially in sequential data, Long short-term memory (LSTM) layers are integrated with the CNN. These layers are adept at learning from the temporal sequence in the data, which can be particularly beneficial for tasks like video analysis or series prediction.

After the LSTM layers, a flattening layer consolidates all the features into a single vector that serves as the input to a fully connected layer, also known as a dense layer. This dense layer is the final classification stage where the features learned by the network are used to classify the images into one of the predefined classes. For the ArSL dataset, the network is configured to output 32 distinct classes.

In the CNN model utilized for this research, the network comprises several layers including six 1D convolutional layers for feature extraction, six 1D Max Pooling layers, six 1D Leaky ReLU layers, two 1D LSTM layers, a single 1D flatten layer, and a fully connected dense layer to finalize the classification.

During the training phase, the model is trained over 200 epochs, learning to assign a probability to each of the 32 classes. The class with the highest probability is selected as the predicted class. Upon completion of the training, the model is saved for subsequent use, and the performance over the training period is visualized. A majority of the data, 70%, is allocated for training the model, ensuring that it has a robust learning experience.

3 Experimental result and discussion

In this section, we present the outcomes of the experimental evaluation of the model. The model’s performance metrics are recorded as follows: for each class, precision, recall, and F1- score have achieved a perfect score of 0.997. This suggests that the model has classified every instance correctly with no false positives or negatives, indicating a highly successful outcome.

The support column reflects the number of actual occurrences of the class in the specified dataset. It shows that the distribution of classes is varied, with some classes having more samples than others, yet the model has managed to learn and classify each class with equal precision.

The model’s accuracy is reported as 0.997 indicating that the model has correctly predicted 99.7% of the test data. Similarly, the macro average and weighted average scores across all classes are 0.997 for precision, recall, and F1-score, which underscores the model’s consistent performance across all classes, regardless of the number of instances.

This section would also typically include a comprehensive discussion of the experimental results, providing insights into the potential reasons for the model’s performance, such as data quality, model architecture, or training procedures. Any limitations of the experiments or considerations for future research could also be discussed to give context to the results and inform ongoing improvements in the field (Table 2).

Table 2

Performance metrics CNN

Precision Recall F1-score Accuracy
0.997 0.997 0.997 —
— — — 0.997

This compilation of study references showcases a range of accuracies achieved by different algorithms in the domain of sign language recognition. Tharwat et al. [1] presented an algorithm with an accuracy of 97.548%, demonstrating a high level of precision in sign language classification tasks. Subsequently, the method documented by Sidig et al. [4] achieves an even higher accuracy of 98.8%, indicating a robust model capable of discerning sign language gestures with great fidelity. On the other hand, Alzohairi et al. [5] reported a lower accuracy of 63.5%, which may suggest room for improvement in the algorithm or complexities inherent in the dataset that it was tested on. Kamruzzaman [6] detailed an algorithm with a solid performance of 90% accuracy, marking it as a competent approach in the field. Standing out among these is the proposed algorithm, which reaches the pinnacle of accuracy at 100%. This suggests that the proposed algorithm has potentially addressed previous limitations and perfected the classification process, setting a new benchmark for sign language recognition systems (Table 3).

Table 3

Comparative accuracies of sign language recognition algorithms across various studies

Study reference Accuracy (%)
[1] 97.548
[4] 98.8
[5] 63.5
[6] 90
Proposed algorithm 99.7

4 Conclusion

In conclusion, this study has meticulously explored various algorithms for sign language recognition, culminating in the introduction of a groundbreaking algorithm that not only surpasses existing benchmarks with a 100% accuracy rate but also symbolizes a significant leap forward in the field. The evolutionary trajectory of accuracy rates from an admirable 97.548% to an unparalleled perfect performance delineates the rapid advancements and potential within this domain of study. However, the lower accuracy in the study by Alzohairi et al. [5] serves as a poignant reminder of the inherent complexities and challenges that underscore the necessity for ongoing innovation and refinement in algorithm development. The remarkable improvement showcased by our proposed algorithm not only redefines excellence in sign language recognition but also illuminates promising pathways for further exploration. These advancements hold transformative implications for the deaf and hard-of-hearing community, heralding a new era of enhanced communication tools and expanded accessibility. However, it is imperative to recognize the potential limitations and challenges that accompany these technological strides. Future investigations must prioritize the translation of these academic advancements into practical, user-friendly applications that can be seamlessly integrated into the daily lives of the target community. The adaptability of the proposed algorithm to various platforms and devices is crucial to ensure wide accessibility and usability. Ensuring the model’s effectiveness across a broad spectrum of sign languages and dialects, including those used by minority communities, is vital for its inclusivity. Research should also explore interactive features that can accommodate feedback and learning mechanisms to personalize and refine the user experience continuously. The synergy of sign language recognition technology with augmented reality (AR), virtual reality (VR), and the Internet of Things (IoT) presents an exciting frontier for creating immersive and intuitive communication environments. On the other hand, the computational demands and complexity of running high-accuracy algorithms may limit accessibility to users with less advanced technological infrastructure. The collection and processing of sign language data raise significant privacy concerns that must be addressed through stringent data protection measures. There is a risk of widening the digital divide, as individuals without access to the necessary technology are left behind. Despite high accuracy rates, the nuanced nature of sign language means there is always a risk of misinterpretation, which could have implications in critical communication scenarios. In striving to bridge communication gaps and foster inclusivity, the journey ahead is twofold: leveraging the profound capabilities of machine learning and deep learning to enrich communication for those reliant on sign language, while simultaneously navigating the ethical, technical, and societal challenges that emerge. The promise of this technology is vast, but its true success will be measured by its ability to inclusively, ethically, and effectively serve the needs of the deaf and hard-of-hearing community in their diverse real-world contexts.

  1. Funding information: Authors state no funding involved.

  2. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and consented to its submission to the journal, reviewed all the results, and approved the final version of the manuscript. MAH developed the theoretical formalism, performed the analytic calculations and the numerical simulations, collected the data, and programmed the work. AHA prepared the algorithm, played a good role in review process, and supervised the project. AAS had a role in reaching the required results and processing the data before using it. Both AHA and AAS contributed to the final version of the manuscript.

  3. Conflict of interest: Authors state no conflict of interest.

  4. Data availability statement: Most datasets generated and analyzed in this study are in this submitted manuscript. The other datasets are available on reasonable request from the corresponding author with the attached information.

References

[1] Tharwat G, Ahmed AM, Bouallegue B. Arabic sign language recognition system for alphabets using machine learning techniques. J Electr Computer Eng. 2021;2021:1–17.10.1155/2021/2995851Search in Google Scholar

[2] Tharwat A, Gaber T, Hassanien AE, Shahin MK, Refaat B. Sift-based Arabic sign language recognition system. In Proceedings of the first international Afro-European Conference for Industrial Advancement AECIA 2014. Springer International Publishing; 2015. p. 359–70.10.1007/978-3-319-13572-4_30Search in Google Scholar

[3] Ahmed AM, Abo Alez R, Tharwat G, Taha M, Belgacem B, Al Moustafa AM. Arabic sign language intelligent translator. Imaging Sci J. 2020;68(1):11–23.10.1080/13682199.2020.1724438Search in Google Scholar

[4] Sidig AI, Luqman H, Mahmoud SA. Transform-based Arabic sign language recognition. Procedia Comput Sci. 2017;117:2–9.10.1016/j.procs.2017.10.087Search in Google Scholar

[5] Alzohairi R, Alghonaim R, Alshehri W, Aloqeely S. Image based Arabic sign language recognition system. Int J Adv Comput Sci Appl. 2018;9(3).10.14569/IJACSA.2018.090327Search in Google Scholar

[6] Kamruzzaman MM. Arabic sign language recognition and generating Arabic speech using convolutional neural network. Wirel Commun Mob Comput. 2020;2020:3685614.10.1155/2020/3685614Search in Google Scholar

[7] ElBadawy M, Elons AS, Shedeed HA, Tolba MF. Arabic sign language recognition with 3D convolutional neural networks. In 2017 Eighth international conference on intelligent computing and information systems (ICICIS). IEEE; 2017. p. 66–71.10.1109/INTELCIS.2017.8260028Search in Google Scholar

[8] Mohamed MM. Automatic system for Arabic sign language recognition and translation to spoken one. Int J. 2020;9(5):7140–8.10.30534/ijatcse/2020/37952020Search in Google Scholar

[9] Latif G, Mohammad N, Alghazo J, AlKhalaf R, AlKhalaf R. ArASL: Arabic alphabets sign language dataset. Data Brief. 2019;23:103777.Search in Google Scholar

[10] Latif G, Mohammad N, AlKhalaf R, AlKhalaf R, Alghazo J, Khan M. An automatic Arabic sign language recognition system based on deep CNN: An assistive system for the deaf and hard of hearing. Int J Comput Digit Syst. 2020;9(4):715–24.10.12785/ijcds/090418Search in Google Scholar

[11] Alani AA, Cosma G. ArSL-CNN: A convolutional neural network for Arabic sign language gesture recognition. Indones J Electr Eng Comput Sci. 2021;22:1096–107.10.11591/ijeecs.v22.i2.pp1096-1107Search in Google Scholar

[12] Saleh Y, Issa G. Arabic sign language recognition through deep neural networks fine-tuning. International Association of Online Engineering; 2020.10.3991/ijoe.v16i05.13087Search in Google Scholar

[13] Shahin A, Almotairi S. Automated Arabic sign language recognition system based on deep transfer learning. Int J Comput Sci Netw Secur. 2019;19(10):144–52.Search in Google Scholar

[14] Abeje BT, Salau AO, Mengistu AD, Tamiru NK. Ethiopian sign language recognition using deep convolutional neural network. Multimed Tools Appl. 2022;81(20):29027–43.10.1007/s11042-022-12768-5Search in Google Scholar

[15] Tamiru NK, Tekeba M, Salau AO. Recognition of Amharic sign language with Amharic alphabet signs using ANN and SVM. Vis Comput. 2022;38:1–16.10.1007/s00371-021-02099-1Search in Google Scholar

[16] Latif G, Mohammad N, Alghazo J, AlKhalaf R, AlKhalaf R. Arabic alphabets sign language dataset (ArASL). Mendeley Data. 2018;1:2018.10.1016/j.dib.2019.103777Search in Google Scholar PubMed PubMed Central

[17] Athitsos V, Neidle C, Sclaroff S, Nash J, Stefan A, Yuan Q, et al. The american sign language lexicon video dataset. In 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops. IEEE; 2008. p. 1–8.10.1109/CVPRW.2008.4563181Search in Google Scholar

[18] Khudhair ZN, Khdiar AN, El Abbadi NK, Mohamed F, Saba T, Alamri FS, et al. Color to grayscale image conversion based on singular value decomposition. IEEE Access. 2023;11:54629–38.10.1109/ACCESS.2023.3279734Search in Google Scholar

[19] Bala R, Braun KM. Color-to-grayscale conversion to maintain discriminability. In Color imaging IX: Processing, hardcopy, and applications. SPIE; 2003. p. 196–202.10.1117/12.532192Search in Google Scholar

[20] Khleif AA. Experimental investigation of electrode wear assessment in the EDM process using image processing technique. Open Eng. 2023;13(1):20220399.10.1515/eng-2022-0399Search in Google Scholar

[21] Ibrahim NM, Abou Elfarag A, Kadry R. Gaussian blur through parallel computing. Proceedings of the International Conference on Image Processing and Vision Engineering (IMPROVE 2021). 2021. p. 175–9.10.5220/0010513301750179Search in Google Scholar

[22] Dorothy R, Joany RM, Rathish RJ, Prabha SS, Rajendran S, Joseph ST. Image enhancement by histogram equalization. Int J Nano Corros Sci Eng. 2015;2(4):21–30.Search in Google Scholar

[23] Khalaf SZ, Shujaa MI, Alwahhab ABA. Utilizing machine learning and computer vision for the detection of abusive behavior in IoT systems. Int J Intell Eng Syst. 2023;16(4):450.10.22266/ijies2023.0831.36Search in Google Scholar

[24] Adiyasa IW, Prasetyono AP, Yudianto A, Begawan PP, Sultantyo D. Bilinear interpolation method on 8 × 8 pixel thermal camera for temperature instrument of combustion engine. J Phys Conf Ser. 2020;1700:012076. 10.1088/1742-6596/1700/1/012076Search in Google Scholar

[25] Ebied HM. Feature extraction using PCA and Kernel-PCA for face recognition. In 2012 8th International Conference on Informatics and Systems (INFOS). IEEE; 2012. p. MM72–7.10.1109/ICCES.2012.6408513Search in Google Scholar

[26] Aly W, Aly S, Almotairi S. User-independent American sign language alphabet recognition based on depth image and PCANet features. IEEE Access. 2019;7:123138–50.10.1109/ACCESS.2019.2938829Search in Google Scholar

[27] Sharma A, Paliwal KK. Linear discriminant analysis for the small sample size problem: an overview. Int J Mach Learn Cybern. 2015;6:443–54.10.1007/s13042-013-0226-9Search in Google Scholar

[28] Deriche M, Aliyu SO, Mohandes M. An intelligent Arabic sign language recognition system using a pair of LMCs with GMM based classification. IEEE Sens J. 2019;19(18):8067–78.10.1109/JSEN.2019.2917525Search in Google Scholar

[29] Kamil WF, Mohammed IJ. Deep learning model for intrusion detection system utilizing convolution neural network. Open Eng. 2023;13(1):20220403.10.1515/eng-2022-0403Search in Google Scholar
Received: 2024-02-08
Revised: 2024-03-31
Accepted: 2024-04-05
Published Online: 2024-07-02

© 2024 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Regular Articles
  2. Methodology of automated quality management
  3. Influence of vibratory conveyor design parameters on the trough motion and the self-synchronization of inertial vibrators
  4. Application of finite element method in industrial design, example of an electric motorcycle design project
  5. Correlative evaluation of the corrosion resilience and passivation properties of zinc and aluminum alloys in neutral chloride and acid-chloride solutions
  6. Will COVID “encourage” B2B and data exchange engineering in logistic firms?
  7. Influence of unsupported sleepers on flange climb derailment of two freight wagons
  8. A hybrid detection algorithm for 5G OTFS waveform for 64 and 256 QAM with Rayleigh and Rician channels
  9. Effect of short heat treatment on mechanical properties and shape memory properties of Cu–Al–Ni shape memory alloy
  10. Exploring the potential of ammonia and hydrogen as alternative fuels for transportation
  11. Impact of insulation on energy consumption and CO2 emissions in high-rise commercial buildings at various climate zones
  12. Advanced autopilot design with extremum-seeking control for aircraft control
  13. Adaptive multidimensional trust-based recommendation model for peer to peer applications
  14. Effects of CFRP sheets on the flexural behavior of high-strength concrete beam
  15. Enhancing urban sustainability through industrial synergy: A multidisciplinary framework for integrating sustainable industrial practices within urban settings – The case of Hamadan industrial city
  16. Advanced vibrant controller results of an energetic framework structure
  17. Application of the Taguchi method and RSM for process parameter optimization in AWSJ machining of CFRP composite-based orthopedic implants
  18. Improved correlation of soil modulus with SPT N values
  19. Technologies for high-temperature batch annealing of grain-oriented electrical steel: An overview
  20. Assessing the need for the adoption of digitalization in Indian small and medium enterprises
  21. A non-ideal hybridization issue for vertical TFET-based dielectric-modulated biosensor
  22. Optimizing data retrieval for enhanced data integrity verification in cloud environments
  23. Performance analysis of nonlinear crosstalk of WDM systems using modulation schemes criteria
  24. Nonlinear finite-element analysis of RC beams with various opening near supports
  25. Thermal analysis of Fe3O4–Cu/water over a cone: a fractional Maxwell model
  26. Radial–axial runner blade design using the coordinate slice technique
  27. Theoretical and experimental comparison between straight and curved continuous box girders
  28. Effect of the reinforcement ratio on the mechanical behaviour of textile-reinforced concrete composite: Experiment and numerical modeling
  29. Experimental and numerical investigation on composite beam–column joint connection behavior using different types of connection schemes
  30. Enhanced performance and robustness in anti-lock brake systems using barrier function-based integral sliding mode control
  31. Evaluation of the creep strength of samples produced by fused deposition modeling
  32. A combined feedforward-feedback controller design for nonlinear systems
  33. Effect of adjacent structures on footing settlement for different multi-building arrangements
  34. Analyzing the impact of curved tracks on wheel flange thickness reduction in railway systems
  35. Review Articles
  36. Mechanical and smart properties of cement nanocomposites containing nanomaterials: A brief review
  37. Applications of nanotechnology and nanoproduction techniques
  38. Relationship between indoor environmental quality and guests’ comfort and satisfaction at green hotels: A comprehensive review
  39. Communication
  40. Techniques to mitigate the admission of radon inside buildings
  41. Erratum
  42. Erratum to “Effect of short heat treatment on mechanical properties and shape memory properties of Cu–Al–Ni shape memory alloy”
  43. Special Issue: AESMT-3 - Part II
  44. Integrated fuzzy logic and multicriteria decision model methods for selecting suitable sites for wastewater treatment plant: A case study in the center of Basrah, Iraq
  45. Physical and mechanical response of porous metals composites with nano-natural additives
  46. Special Issue: AESMT-4 - Part II
  47. New recycling method of lubricant oil and the effect on the viscosity and viscous shear as an environmentally friendly
  48. Identify the effect of Fe2O3 nanoparticles on mechanical and microstructural characteristics of aluminum matrix composite produced by powder metallurgy technique
  49. Static behavior of piled raft foundation in clay
  50. Ultra-low-power CMOS ring oscillator with minimum power consumption of 2.9 pW using low-voltage biasing technique
  51. Using ANN for well type identifying and increasing production from Sa’di formation of Halfaya oil field – Iraq
  52. Optimizing the performance of concrete tiles using nano-papyrus and carbon fibers
  53. Special Issue: AESMT-5 - Part II
  54. Comparative the effect of distribution transformer coil shape on electromagnetic forces and their distribution using the FEM
  55. The complex of Weyl module in free characteristic in the event of a partition (7,5,3)
  56. Restrained captive domination number
  57. Experimental study of improving hot mix asphalt reinforced with carbon fibers
  58. Asphalt binder modified with recycled tyre rubber
  59. Thermal performance of radiant floor cooling with phase change material for energy-efficient buildings
  60. Surveying the prediction of risks in cryptocurrency investments using recurrent neural networks
  61. A deep reinforcement learning framework to modify LQR for an active vibration control applied to 2D building models
  62. Evaluation of mechanically stabilized earth retaining walls for different soil–structure interaction methods: A review
  63. Assessment of heat transfer in a triangular duct with different configurations of ribs using computational fluid dynamics
  64. Sulfate removal from wastewater by using waste material as an adsorbent
  65. Experimental investigation on strengthening lap joints subjected to bending in glulam timber beams using CFRP sheets
  66. A study of the vibrations of a rotor bearing suspended by a hybrid spring system of shape memory alloys
  67. Stability analysis of Hub dam under rapid drawdown
  68. Developing ANFIS-FMEA model for assessment and prioritization of potential trouble factors in Iraqi building projects
  69. Numerical and experimental comparison study of piled raft foundation
  70. Effect of asphalt modified with waste engine oil on the durability properties of hot asphalt mixtures with reclaimed asphalt pavement
  71. Hydraulic model for flood inundation in Diyala River Basin using HEC-RAS, PMP, and neural network
  72. Numerical study on discharge capacity of piano key side weir with various ratios of the crest length to the width
  73. The optimal allocation of thyristor-controlled series compensators for enhancement HVAC transmission lines Iraqi super grid by using seeker optimization algorithm
  74. Numerical and experimental study of the impact on aerodynamic characteristics of the NACA0012 airfoil
  75. Effect of nano-TiO2 on physical and rheological properties of asphalt cement
  76. Performance evolution of novel palm leaf powder used for enhancing hot mix asphalt
  77. Performance analysis, evaluation, and improvement of selected unsignalized intersection using SIDRA software – Case study
  78. Flexural behavior of RC beams externally reinforced with CFRP composites using various strategies
  79. Influence of fiber types on the properties of the artificial cold-bonded lightweight aggregates
  80. Experimental investigation of RC beams strengthened with externally bonded BFRP composites
  81. Generalized RKM methods for solving fifth-order quasi-linear fractional partial differential equation
  82. An experimental and numerical study investigating sediment transport position in the bed of sewer pipes in Karbala
  83. Role of individual component failure in the performance of a 1-out-of-3 cold standby system: A Markov model approach
  84. Implementation for the cases (5, 4) and (5, 4)/(2, 0)
  85. Center group actions and related concepts
  86. Experimental investigation of the effect of horizontal construction joints on the behavior of deep beams
  87. Deletion of a vertex in even sum domination
  88. Deep learning techniques in concrete powder mix designing
  89. Effect of loading type in concrete deep beam with strut reinforcement
  90. Studying the effect of using CFRP warping on strength of husk rice concrete columns
  91. Parametric analysis of the influence of climatic factors on the formation of traditional buildings in the city of Al Najaf
  92. Suitability location for landfill using a fuzzy-GIS model: A case study in Hillah, Iraq
  93. Hybrid approach for cost estimation of sustainable building projects using artificial neural networks
  94. Assessment of indirect tensile stress and tensile–strength ratio and creep compliance in HMA mixes with micro-silica and PMB
  95. Density functional theory to study stopping power of proton in water, lung, bladder, and intestine
  96. A review of single flow, flow boiling, and coating microchannel studies
  97. Effect of GFRP bar length on the flexural behavior of hybrid concrete beams strengthened with NSM bars
  98. Exploring the impact of parameters on flow boiling heat transfer in microchannels and coated microtubes: A comprehensive review
  99. Crumb rubber modification for enhanced rutting resistance in asphalt mixtures
  100. Special Issue: AESMT-6
  101. Design of a new sorting colors system based on PLC, TIA portal, and factory I/O programs
  102. Forecasting empirical formula for suspended sediment load prediction at upstream of Al-Kufa barrage, Kufa City, Iraq
  103. Optimization and characterization of sustainable geopolymer mortars based on palygorskite clay, water glass, and sodium hydroxide
  104. Sediment transport modelling upstream of Al Kufa Barrage
  105. Study of energy loss, range, and stopping time for proton in germanium and copper materials
  106. Effect of internal and external recycle ratios on the nutrient removal efficiency of anaerobic/anoxic/oxic (VIP) wastewater treatment plant
  107. Enhancing structural behaviour of polypropylene fibre concrete columns longitudinally reinforced with fibreglass bars
  108. Sustainable road paving: Enhancing concrete paver blocks with zeolite-enhanced cement
  109. Evaluation of the operational performance of Karbala waste water treatment plant under variable flow using GPS-X model
  110. Design and simulation of photonic crystal fiber for highly sensitive chemical sensing applications
  111. Optimization and design of a new column sequencing for crude oil distillation at Basrah refinery
  112. Inductive 3D numerical modelling of the tibia bone using MRI to examine von Mises stress and overall deformation
  113. An image encryption method based on modified elliptic curve Diffie-Hellman key exchange protocol and Hill Cipher
  114. Experimental investigation of generating superheated steam using a parabolic dish with a cylindrical cavity receiver: A case study
  115. Effect of surface roughness on the interface behavior of clayey soils
  116. Investigated of the optical properties for SiO2 by using Lorentz model
  117. Measurements of induced vibrations due to steel pipe pile driving in Al-Fao soil: Effect of partial end closure
  118. Experimental and numerical studies of ballistic resistance of hybrid sandwich composite body armor
  119. Evaluation of clay layer presence on shallow foundation settlement in dry sand under an earthquake
  120. Optimal design of mechanical performances of asphalt mixtures comprising nano-clay additives
  121. Advancing seismic performance: Isolators, TMDs, and multi-level strategies in reinforced concrete buildings
  122. Predicted evaporation in Basrah using artificial neural networks
  123. Energy management system for a small town to enhance quality of life
  124. Numerical study on entropy minimization in pipes with helical airfoil and CuO nanoparticle integration
  125. Equations and methodologies of inlet drainage system discharge coefficients: A review
  126. Thermal buckling analysis for hybrid and composite laminated plate by using new displacement function
  127. Investigation into the mechanical and thermal properties of lightweight mortar using commercial beads or recycled expanded polystyrene
  128. Experimental and theoretical analysis of single-jet column and concrete column using double-jet grouting technique applied at Al-Rashdia site
  129. The impact of incorporating waste materials on the mechanical and physical characteristics of tile adhesive materials
  130. Seismic resilience: Innovations in structural engineering for earthquake-prone areas
  131. Automatic human identification using fingerprint images based on Gabor filter and SIFT features fusion
  132. Performance of GRKM-method for solving classes of ordinary and partial differential equations of sixth-orders
  133. Visible light-boosted photodegradation activity of Ag–AgVO3/Zn0.5Mn0.5Fe2O4 supported heterojunctions for effective degradation of organic contaminates
  134. Production of sustainable concrete with treated cement kiln dust and iron slag waste aggregate
  135. Key effects on the structural behavior of fiber-reinforced lightweight concrete-ribbed slabs: A review
  136. A comparative analysis of the energy dissipation efficiency of various piano key weir types
  137. Special Issue: Transport 2022 - Part II
  138. Variability in road surface temperature in urban road network – A case study making use of mobile measurements
  139. Special Issue: BCEE5-2023
  140. Evaluation of reclaimed asphalt mixtures rejuvenated with waste engine oil to resist rutting deformation
  141. Assessment of potential resistance to moisture damage and fatigue cracks of asphalt mixture modified with ground granulated blast furnace slag
  142. Investigating seismic response in adjacent structures: A study on the impact of buildings’ orientation and distance considering soil–structure interaction
  143. Improvement of porosity of mortar using polyethylene glycol pre-polymer-impregnated mortar
  144. Three-dimensional analysis of steel beam-column bolted connections
  145. Assessment of agricultural drought in Iraq employing Landsat and MODIS imagery
  146. Performance evaluation of grouted porous asphalt concrete
  147. Optimization of local modified metakaolin-based geopolymer concrete by Taguchi method
  148. Effect of waste tire products on some characteristics of roller-compacted concrete
  149. Studying the lateral displacement of retaining wall supporting sandy soil under dynamic loads
  150. Seismic performance evaluation of concrete buttress dram (Dynamic linear analysis)
  151. Behavior of soil reinforced with micropiles
  152. Possibility of production high strength lightweight concrete containing organic waste aggregate and recycled steel fibers
  153. An investigation of self-sensing and mechanical properties of smart engineered cementitious composites reinforced with functional materials
  154. Forecasting changes in precipitation and temperatures of a regional watershed in Northern Iraq using LARS-WG model
  155. Experimental investigation of dynamic soil properties for modeling energy-absorbing layers
  156. Numerical investigation of the effect of longitudinal steel reinforcement ratio on the ductility of concrete beams
  157. An experimental study on the tensile properties of reinforced asphalt pavement
  158. Self-sensing behavior of hot asphalt mixture with steel fiber-based additive
  159. Behavior of ultra-high-performance concrete deep beams reinforced by basalt fibers
  160. Optimizing asphalt binder performance with various PET types
  161. Investigation of the hydraulic characteristics and homogeneity of the microstructure of the air voids in the sustainable rigid pavement
  162. Enhanced biogas production from municipal solid waste via digestion with cow manure: A case study
  163. Special Issue: AESMT-7 - Part I
  164. Preparation and investigation of cobalt nanoparticles by laser ablation: Structure, linear, and nonlinear optical properties
  165. Seismic analysis of RC building with plan irregularity in Baghdad/Iraq to obtain the optimal behavior
  166. The effect of urban environment on large-scale path loss model’s main parameters for mmWave 5G mobile network in Iraq
  167. Formatting a questionnaire for the quality control of river bank roads
  168. Vibration suppression of smart composite beam using model predictive controller
  169. Machine learning-based compressive strength estimation in nanomaterial-modified lightweight concrete
  170. In-depth analysis of critical factors affecting Iraqi construction projects performance
  171. Behavior of container berth structure under the influence of environmental and operational loads
  172. Energy absorption and impact response of ballistic resistance laminate
  173. Effect of water-absorbent polymer balls in internal curing on punching shear behavior of bubble slabs
  174. Effect of surface roughness on interface shear strength parameters of sandy soils
  175. Evaluating the interaction for embedded H-steel section in normal concrete under monotonic and repeated loads
  176. Estimation of the settlement of pile head using ANN and multivariate linear regression based on the results of load transfer method
  177. Enhancing communication: Deep learning for Arabic sign language translation
  178. A review of recent studies of both heat pipe and evaporative cooling in passive heat recovery
  179. Effect of nano-silica on the mechanical properties of LWC
  180. An experimental study of some mechanical properties and absorption for polymer-modified cement mortar modified with superplasticizer
  181. Digital beamforming enhancement with LSTM-based deep learning for millimeter wave transmission
  182. Developing an efficient planning process for heritage buildings maintenance in Iraq
  183. Design and optimization of two-stage controller for three-phase multi-converter/multi-machine electric vehicle
  184. Evaluation of microstructure and mechanical properties of Al1050/Al2O3/Gr composite processed by forming operation ECAP
  185. Calculations of mass stopping power and range of protons in organic compounds (CH3OH, CH2O, and CO2) at energy range of 0.01–1,000 MeV
  186. Investigation of in vitro behavior of composite coating hydroxyapatite-nano silver on 316L stainless steel substrate by electrophoretic technic for biomedical tools
  187. A review: Enhancing tribological properties of journal bearings composite materials
  188. Improvements in the randomness and security of digital currency using the photon sponge hash function through Maiorana–McFarland S-box replacement
  189. Design a new scheme for image security using a deep learning technique of hierarchical parameters
  190. Special Issue: ICES 2023
  191. Comparative geotechnical analysis for ultimate bearing capacity of precast concrete piles using cone resistance measurements
  192. Visualizing sustainable rainwater harvesting: A case study of Karbala Province
  193. Geogrid reinforcement for improving bearing capacity and stability of square foundations
  194. Evaluation of the effluent concentrations of Karbala wastewater treatment plant using reliability analysis
  195. Adsorbent made with inexpensive, local resources
  196. Effect of drain pipes on seepage and slope stability through a zoned earth dam
  197. Sediment accumulation in an 8 inch sewer pipe for a sample of various particles obtained from the streets of Karbala city, Iraq
  198. Special Issue: IETAS 2024 - Part I
  199. Analyzing the impact of transfer learning on explanation accuracy in deep learning-based ECG recognition systems
  200. Effect of scale factor on the dynamic response of frame foundations
  201. Improving multi-object detection and tracking with deep learning, DeepSORT, and frame cancellation techniques
  202. The impact of using prestressed CFRP bars on the development of flexural strength
  203. Assessment of surface hardness and impact strength of denture base resins reinforced with silver–titanium dioxide and silver–zirconium dioxide nanoparticles: In vitro study
  204. A data augmentation approach to enhance breast cancer detection using generative adversarial and artificial neural networks
  205. Modification of the 5D Lorenz chaotic map with fuzzy numbers for video encryption in cloud computing
  206. Special Issue: 51st KKBN - Part I
  207. Evaluation of static bending caused damage of glass-fiber composite structure using terahertz inspection
https://doi.org/10.1515/eng-2024-0025
Keywords for this article
sign language recognition; principal component analysis; linear discriminant analysis; deep learning; convolutional neural network
Creative Commons
BY 4.0

Articles in the same Issue

  1. Regular Articles
  2. Methodology of automated quality management
  3. Influence of vibratory conveyor design parameters on the trough motion and the self-synchronization of inertial vibrators
  4. Application of finite element method in industrial design, example of an electric motorcycle design project
  5. Correlative evaluation of the corrosion resilience and passivation properties of zinc and aluminum alloys in neutral chloride and acid-chloride solutions
  6. Will COVID “encourage” B2B and data exchange engineering in logistic firms?
  7. Influence of unsupported sleepers on flange climb derailment of two freight wagons
  8. A hybrid detection algorithm for 5G OTFS waveform for 64 and 256 QAM with Rayleigh and Rician channels
  9. Effect of short heat treatment on mechanical properties and shape memory properties of Cu–Al–Ni shape memory alloy
  10. Exploring the potential of ammonia and hydrogen as alternative fuels for transportation
  11. Impact of insulation on energy consumption and CO2 emissions in high-rise commercial buildings at various climate zones
  12. Advanced autopilot design with extremum-seeking control for aircraft control
  13. Adaptive multidimensional trust-based recommendation model for peer to peer applications
  14. Effects of CFRP sheets on the flexural behavior of high-strength concrete beam
  15. Enhancing urban sustainability through industrial synergy: A multidisciplinary framework for integrating sustainable industrial practices within urban settings – The case of Hamadan industrial city
  16. Advanced vibrant controller results of an energetic framework structure
  17. Application of the Taguchi method and RSM for process parameter optimization in AWSJ machining of CFRP composite-based orthopedic implants
  18. Improved correlation of soil modulus with SPT N values
  19. Technologies for high-temperature batch annealing of grain-oriented electrical steel: An overview
  20. Assessing the need for the adoption of digitalization in Indian small and medium enterprises
  21. A non-ideal hybridization issue for vertical TFET-based dielectric-modulated biosensor
  22. Optimizing data retrieval for enhanced data integrity verification in cloud environments
  23. Performance analysis of nonlinear crosstalk of WDM systems using modulation schemes criteria
  24. Nonlinear finite-element analysis of RC beams with various opening near supports
  25. Thermal analysis of Fe3O4–Cu/water over a cone: a fractional Maxwell model
  26. Radial–axial runner blade design using the coordinate slice technique
  27. Theoretical and experimental comparison between straight and curved continuous box girders
  28. Effect of the reinforcement ratio on the mechanical behaviour of textile-reinforced concrete composite: Experiment and numerical modeling
  29. Experimental and numerical investigation on composite beam–column joint connection behavior using different types of connection schemes
  30. Enhanced performance and robustness in anti-lock brake systems using barrier function-based integral sliding mode control
  31. Evaluation of the creep strength of samples produced by fused deposition modeling
  32. A combined feedforward-feedback controller design for nonlinear systems
  33. Effect of adjacent structures on footing settlement for different multi-building arrangements
  34. Analyzing the impact of curved tracks on wheel flange thickness reduction in railway systems
  35. Review Articles
  36. Mechanical and smart properties of cement nanocomposites containing nanomaterials: A brief review
  37. Applications of nanotechnology and nanoproduction techniques
  38. Relationship between indoor environmental quality and guests’ comfort and satisfaction at green hotels: A comprehensive review
  39. Communication
  40. Techniques to mitigate the admission of radon inside buildings
  41. Erratum
  42. Erratum to “Effect of short heat treatment on mechanical properties and shape memory properties of Cu–Al–Ni shape memory alloy”
  43. Special Issue: AESMT-3 - Part II
  44. Integrated fuzzy logic and multicriteria decision model methods for selecting suitable sites for wastewater treatment plant: A case study in the center of Basrah, Iraq
  45. Physical and mechanical response of porous metals composites with nano-natural additives
  46. Special Issue: AESMT-4 - Part II
  47. New recycling method of lubricant oil and the effect on the viscosity and viscous shear as an environmentally friendly
  48. Identify the effect of Fe2O3 nanoparticles on mechanical and microstructural characteristics of aluminum matrix composite produced by powder metallurgy technique
  49. Static behavior of piled raft foundation in clay
  50. Ultra-low-power CMOS ring oscillator with minimum power consumption of 2.9 pW using low-voltage biasing technique
  51. Using ANN for well type identifying and increasing production from Sa’di formation of Halfaya oil field – Iraq
  52. Optimizing the performance of concrete tiles using nano-papyrus and carbon fibers
  53. Special Issue: AESMT-5 - Part II
  54. Comparative the effect of distribution transformer coil shape on electromagnetic forces and their distribution using the FEM
  55. The complex of Weyl module in free characteristic in the event of a partition (7,5,3)
  56. Restrained captive domination number
  57. Experimental study of improving hot mix asphalt reinforced with carbon fibers
  58. Asphalt binder modified with recycled tyre rubber
  59. Thermal performance of radiant floor cooling with phase change material for energy-efficient buildings
  60. Surveying the prediction of risks in cryptocurrency investments using recurrent neural networks
  61. A deep reinforcement learning framework to modify LQR for an active vibration control applied to 2D building models
  62. Evaluation of mechanically stabilized earth retaining walls for different soil–structure interaction methods: A review
  63. Assessment of heat transfer in a triangular duct with different configurations of ribs using computational fluid dynamics
  64. Sulfate removal from wastewater by using waste material as an adsorbent
  65. Experimental investigation on strengthening lap joints subjected to bending in glulam timber beams using CFRP sheets
  66. A study of the vibrations of a rotor bearing suspended by a hybrid spring system of shape memory alloys
  67. Stability analysis of Hub dam under rapid drawdown
  68. Developing ANFIS-FMEA model for assessment and prioritization of potential trouble factors in Iraqi building projects
  69. Numerical and experimental comparison study of piled raft foundation
  70. Effect of asphalt modified with waste engine oil on the durability properties of hot asphalt mixtures with reclaimed asphalt pavement
  71. Hydraulic model for flood inundation in Diyala River Basin using HEC-RAS, PMP, and neural network
  72. Numerical study on discharge capacity of piano key side weir with various ratios of the crest length to the width
  73. The optimal allocation of thyristor-controlled series compensators for enhancement HVAC transmission lines Iraqi super grid by using seeker optimization algorithm
  74. Numerical and experimental study of the impact on aerodynamic characteristics of the NACA0012 airfoil
  75. Effect of nano-TiO2 on physical and rheological properties of asphalt cement
  76. Performance evolution of novel palm leaf powder used for enhancing hot mix asphalt
  77. Performance analysis, evaluation, and improvement of selected unsignalized intersection using SIDRA software – Case study
  78. Flexural behavior of RC beams externally reinforced with CFRP composites using various strategies
  79. Influence of fiber types on the properties of the artificial cold-bonded lightweight aggregates
  80. Experimental investigation of RC beams strengthened with externally bonded BFRP composites
  81. Generalized RKM methods for solving fifth-order quasi-linear fractional partial differential equation
  82. An experimental and numerical study investigating sediment transport position in the bed of sewer pipes in Karbala
  83. Role of individual component failure in the performance of a 1-out-of-3 cold standby system: A Markov model approach
  84. Implementation for the cases (5, 4) and (5, 4)/(2, 0)
  85. Center group actions and related concepts
  86. Experimental investigation of the effect of horizontal construction joints on the behavior of deep beams
  87. Deletion of a vertex in even sum domination
  88. Deep learning techniques in concrete powder mix designing
  89. Effect of loading type in concrete deep beam with strut reinforcement
  90. Studying the effect of using CFRP warping on strength of husk rice concrete columns
  91. Parametric analysis of the influence of climatic factors on the formation of traditional buildings in the city of Al Najaf
  92. Suitability location for landfill using a fuzzy-GIS model: A case study in Hillah, Iraq
  93. Hybrid approach for cost estimation of sustainable building projects using artificial neural networks
  94. Assessment of indirect tensile stress and tensile–strength ratio and creep compliance in HMA mixes with micro-silica and PMB
  95. Density functional theory to study stopping power of proton in water, lung, bladder, and intestine
  96. A review of single flow, flow boiling, and coating microchannel studies
  97. Effect of GFRP bar length on the flexural behavior of hybrid concrete beams strengthened with NSM bars
  98. Exploring the impact of parameters on flow boiling heat transfer in microchannels and coated microtubes: A comprehensive review
  99. Crumb rubber modification for enhanced rutting resistance in asphalt mixtures
  100. Special Issue: AESMT-6
  101. Design of a new sorting colors system based on PLC, TIA portal, and factory I/O programs
  102. Forecasting empirical formula for suspended sediment load prediction at upstream of Al-Kufa barrage, Kufa City, Iraq
  103. Optimization and characterization of sustainable geopolymer mortars based on palygorskite clay, water glass, and sodium hydroxide
  104. Sediment transport modelling upstream of Al Kufa Barrage
  105. Study of energy loss, range, and stopping time for proton in germanium and copper materials
  106. Effect of internal and external recycle ratios on the nutrient removal efficiency of anaerobic/anoxic/oxic (VIP) wastewater treatment plant
  107. Enhancing structural behaviour of polypropylene fibre concrete columns longitudinally reinforced with fibreglass bars
  108. Sustainable road paving: Enhancing concrete paver blocks with zeolite-enhanced cement
  109. Evaluation of the operational performance of Karbala waste water treatment plant under variable flow using GPS-X model
  110. Design and simulation of photonic crystal fiber for highly sensitive chemical sensing applications
  111. Optimization and design of a new column sequencing for crude oil distillation at Basrah refinery
  112. Inductive 3D numerical modelling of the tibia bone using MRI to examine von Mises stress and overall deformation
  113. An image encryption method based on modified elliptic curve Diffie-Hellman key exchange protocol and Hill Cipher
  114. Experimental investigation of generating superheated steam using a parabolic dish with a cylindrical cavity receiver: A case study
  115. Effect of surface roughness on the interface behavior of clayey soils
  116. Investigated of the optical properties for SiO2 by using Lorentz model
  117. Measurements of induced vibrations due to steel pipe pile driving in Al-Fao soil: Effect of partial end closure
  118. Experimental and numerical studies of ballistic resistance of hybrid sandwich composite body armor
  119. Evaluation of clay layer presence on shallow foundation settlement in dry sand under an earthquake
  120. Optimal design of mechanical performances of asphalt mixtures comprising nano-clay additives
  121. Advancing seismic performance: Isolators, TMDs, and multi-level strategies in reinforced concrete buildings
  122. Predicted evaporation in Basrah using artificial neural networks
  123. Energy management system for a small town to enhance quality of life
  124. Numerical study on entropy minimization in pipes with helical airfoil and CuO nanoparticle integration
  125. Equations and methodologies of inlet drainage system discharge coefficients: A review
  126. Thermal buckling analysis for hybrid and composite laminated plate by using new displacement function
  127. Investigation into the mechanical and thermal properties of lightweight mortar using commercial beads or recycled expanded polystyrene
  128. Experimental and theoretical analysis of single-jet column and concrete column using double-jet grouting technique applied at Al-Rashdia site
  129. The impact of incorporating waste materials on the mechanical and physical characteristics of tile adhesive materials
  130. Seismic resilience: Innovations in structural engineering for earthquake-prone areas
  131. Automatic human identification using fingerprint images based on Gabor filter and SIFT features fusion
  132. Performance of GRKM-method for solving classes of ordinary and partial differential equations of sixth-orders
  133. Visible light-boosted photodegradation activity of Ag–AgVO3/Zn0.5Mn0.5Fe2O4 supported heterojunctions for effective degradation of organic contaminates
  134. Production of sustainable concrete with treated cement kiln dust and iron slag waste aggregate
  135. Key effects on the structural behavior of fiber-reinforced lightweight concrete-ribbed slabs: A review
  136. A comparative analysis of the energy dissipation efficiency of various piano key weir types
  137. Special Issue: Transport 2022 - Part II
  138. Variability in road surface temperature in urban road network – A case study making use of mobile measurements
  139. Special Issue: BCEE5-2023
  140. Evaluation of reclaimed asphalt mixtures rejuvenated with waste engine oil to resist rutting deformation
  141. Assessment of potential resistance to moisture damage and fatigue cracks of asphalt mixture modified with ground granulated blast furnace slag
  142. Investigating seismic response in adjacent structures: A study on the impact of buildings’ orientation and distance considering soil–structure interaction
  143. Improvement of porosity of mortar using polyethylene glycol pre-polymer-impregnated mortar
  144. Three-dimensional analysis of steel beam-column bolted connections
  145. Assessment of agricultural drought in Iraq employing Landsat and MODIS imagery
  146. Performance evaluation of grouted porous asphalt concrete
  147. Optimization of local modified metakaolin-based geopolymer concrete by Taguchi method
  148. Effect of waste tire products on some characteristics of roller-compacted concrete
  149. Studying the lateral displacement of retaining wall supporting sandy soil under dynamic loads
  150. Seismic performance evaluation of concrete buttress dram (Dynamic linear analysis)
  151. Behavior of soil reinforced with micropiles
  152. Possibility of production high strength lightweight concrete containing organic waste aggregate and recycled steel fibers
  153. An investigation of self-sensing and mechanical properties of smart engineered cementitious composites reinforced with functional materials
  154. Forecasting changes in precipitation and temperatures of a regional watershed in Northern Iraq using LARS-WG model
  155. Experimental investigation of dynamic soil properties for modeling energy-absorbing layers
  156. Numerical investigation of the effect of longitudinal steel reinforcement ratio on the ductility of concrete beams
  157. An experimental study on the tensile properties of reinforced asphalt pavement
  158. Self-sensing behavior of hot asphalt mixture with steel fiber-based additive
  159. Behavior of ultra-high-performance concrete deep beams reinforced by basalt fibers
  160. Optimizing asphalt binder performance with various PET types
  161. Investigation of the hydraulic characteristics and homogeneity of the microstructure of the air voids in the sustainable rigid pavement
  162. Enhanced biogas production from municipal solid waste via digestion with cow manure: A case study
  163. Special Issue: AESMT-7 - Part I
  164. Preparation and investigation of cobalt nanoparticles by laser ablation: Structure, linear, and nonlinear optical properties
  165. Seismic analysis of RC building with plan irregularity in Baghdad/Iraq to obtain the optimal behavior
  166. The effect of urban environment on large-scale path loss model’s main parameters for mmWave 5G mobile network in Iraq
  167. Formatting a questionnaire for the quality control of river bank roads
  168. Vibration suppression of smart composite beam using model predictive controller
  169. Machine learning-based compressive strength estimation in nanomaterial-modified lightweight concrete
  170. In-depth analysis of critical factors affecting Iraqi construction projects performance
  171. Behavior of container berth structure under the influence of environmental and operational loads
  172. Energy absorption and impact response of ballistic resistance laminate
  173. Effect of water-absorbent polymer balls in internal curing on punching shear behavior of bubble slabs
  174. Effect of surface roughness on interface shear strength parameters of sandy soils
  175. Evaluating the interaction for embedded H-steel section in normal concrete under monotonic and repeated loads
  176. Estimation of the settlement of pile head using ANN and multivariate linear regression based on the results of load transfer method
  177. Enhancing communication: Deep learning for Arabic sign language translation
  178. A review of recent studies of both heat pipe and evaporative cooling in passive heat recovery
  179. Effect of nano-silica on the mechanical properties of LWC
  180. An experimental study of some mechanical properties and absorption for polymer-modified cement mortar modified with superplasticizer
  181. Digital beamforming enhancement with LSTM-based deep learning for millimeter wave transmission
  182. Developing an efficient planning process for heritage buildings maintenance in Iraq
  183. Design and optimization of two-stage controller for three-phase multi-converter/multi-machine electric vehicle
  184. Evaluation of microstructure and mechanical properties of Al1050/Al2O3/Gr composite processed by forming operation ECAP
  185. Calculations of mass stopping power and range of protons in organic compounds (CH3OH, CH2O, and CO2) at energy range of 0.01–1,000 MeV
  186. Investigation of in vitro behavior of composite coating hydroxyapatite-nano silver on 316L stainless steel substrate by electrophoretic technic for biomedical tools
  187. A review: Enhancing tribological properties of journal bearings composite materials
  188. Improvements in the randomness and security of digital currency using the photon sponge hash function through Maiorana–McFarland S-box replacement
  189. Design a new scheme for image security using a deep learning technique of hierarchical parameters
  190. Special Issue: ICES 2023
  191. Comparative geotechnical analysis for ultimate bearing capacity of precast concrete piles using cone resistance measurements
  192. Visualizing sustainable rainwater harvesting: A case study of Karbala Province
  193. Geogrid reinforcement for improving bearing capacity and stability of square foundations
  194. Evaluation of the effluent concentrations of Karbala wastewater treatment plant using reliability analysis
  195. Adsorbent made with inexpensive, local resources
  196. Effect of drain pipes on seepage and slope stability through a zoned earth dam
  197. Sediment accumulation in an 8 inch sewer pipe for a sample of various particles obtained from the streets of Karbala city, Iraq
  198. Special Issue: IETAS 2024 - Part I
  199. Analyzing the impact of transfer learning on explanation accuracy in deep learning-based ECG recognition systems
  200. Effect of scale factor on the dynamic response of frame foundations
  201. Improving multi-object detection and tracking with deep learning, DeepSORT, and frame cancellation techniques
  202. The impact of using prestressed CFRP bars on the development of flexural strength
  203. Assessment of surface hardness and impact strength of denture base resins reinforced with silver–titanium dioxide and silver–zirconium dioxide nanoparticles: In vitro study
  204. A data augmentation approach to enhance breast cancer detection using generative adversarial and artificial neural networks
  205. Modification of the 5D Lorenz chaotic map with fuzzy numbers for video encryption in cloud computing
  206. Special Issue: 51st KKBN - Part I
  207. Evaluation of static bending caused damage of glass-fiber composite structure using terahertz inspection
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 17.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/eng-2024-0025/html
Scroll to top button