Iman Maaly | University of Khartoum (original) (raw)

Papers by Iman Maaly

University of Khartoum Engineering Journal

In this work, a Multiple Low Latency Queuing scheduling mechanism model is developed to improve t... more In this work, a Multiple Low Latency Queuing scheduling mechanism model is developed to improve the QoS performance for real time and critical mission data traffic in LTE mobile networks. The main objective of this model is to achieve minimum delay and improve the QoS for real time applications (like Live Video and Voice over LTE). In addition, issues likestarvation of lower priority queues and bandwidth allocation are addressed. The model is composed of four components, first, classifier to classify the incoming traffic in router interface. Second, four Class Based Weighted Fair Queues (CBWFQ) scheduling mechanisms, with activation of strict priority feature in the first two queues. Third, two separate rate limiters (policers), one for each strict priority queue. Two scenarios are designed and simulated using Optimized Network Engineering Tool (OPNET). The results show that, in the case of Multiple Low Latency Queuing scheduling mechanism model, the real time traffic suffers less d...

2019 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE)

this study is a comparison study with the purpose to propose an approach for selecting the best c... more this study is a comparison study with the purpose to propose an approach for selecting the best classifier for diagnoses of atrial fibrillation (AF) in coronary heartbeats. The Physionet Computing in Cardiology Challenge 2017 turned into used as the data source for this study. Automatic ECG processing consists essentially of the detection and location of the signal characteristic points and is an important tool in the management of cardiac diseases. The most relevant task is the detection of the QRS complex after which a complete analysis and delineation of each beat can be obtained. In the preprocessing stage, the Discrete Wavelet Transform (DWT) is used for removing noise and tuning to the morphological characteristics of the waveform features. For feature extraction, a set of features that consists of both morphological and temporal features is extracted using DWT. A comparison study was conducted between five classifiers (Decision trees, Random forest, AdaBoost ensemble classifier, support vector machine (SVM) and K-nearest neighbor Algorithm (KNN)) to know which give the best diagnoses for each type of Arrhythmia. In this study we used four classes of coronary heart beats atrial fibrillation, normal, other rhythms or noise. Results show that the AdaBoost classifier gives 100 % Accuracy scores for all types of Arrhythmia in the training set. The AdaBoost algorithm obtained a mean improvement report for all classes in testing set (97.3% in Area under curve accuracy (AUC), 94.7% in classifier accuracy (CA), 96.7% in sensitivity (Recall), and of 98 % in positive predictive value (Precision)). Keywords: Electrocardiogram

An Institutional Repository (IR) is a set of services that a university offers to its community m... more An Institutional Repository (IR) is a set of services that a university offers to its community members for the management and dissemination of digital materials created by its researchers and staff members. IR establishment is a challenge for information professionals because the main goal for implementing IRs is to provide Open Access (OA) to institution research output. The growth of open access institutional repositories has been very remarkable in the developed countries. However, in most of the developing countries like Sudan awareness of the concept of open access and IR is very low. Since its foundation, the Sudanese Research and Education Network (SudREN) has successfully implemented several library projects. The most successful one was a project funded by UNESCO in 2013. The main objective of that project was to implement Library Management Systems (LMS) and Institutional Repositories (IR) for three Sudanese universities based on open source Software Systems. Project compo...

WIT Transactions on Information and Communication Technologies, 2002

A Hidden Markov Model (HMM) recognition system is implemented for Arabic Phonemes as units of rec... more A Hidden Markov Model (HMM) recognition system is implemented for Arabic Phonemes as units of recognition. An important result of this system is the confision encountered between some phonemes of Arabic (e.g., /h/ (Ha), /?/ (Hamza)), i.e., the recognize could not distinguish between them. New parameters, which are based on a new classification of Arabic phonemes, are added at a higher level of the system for resolving this acoustic confusability and improving the recognition accuracy. These new parameters, which are based on the shape of the tongue and the place of articulation along the vocal tract, are called the Emphatic/nonemphatic, Root and Hamz parameters. The “Emphatic/nonemphatic” parameter gives a performance of 44%, the “Root” parameter gives a performance of 40%, while the “Hamz” parameter could resolve confusibility encountered between the phonemes /h/ and /?/ with a performance of 90%. These new parameters may need accurate estimates of the distances along the vocal tra...

WIT Transactions on Information and Communication Technologies, 1970

Journal of Biomedical Engineering and Medical Imaging, 2020

Automation of human alertness identification has been widely investigated in recent decades. Many... more Automation of human alertness identification has been widely investigated in recent decades. Many applications can benefit from automatic alertness state identification, such as driver fatigue detection, monotonous task workers' vigilance detection and sleep studies in the medical field. Many researchers have tried to exploit different types of behavioural aspects in vigilance detection, such as eye movement, head position and facial expression. On the other hand, some biomedical signals like ECG, EEG and heart rhythm are also exploited; however, there is a consensus of the superiority of EEG signal in alertness classification due to its close relation with different human vigilant states. In this paper, we propose an automatic method for vigilance detection using a single EEG channel along with sparse representation and dictionary learning. We used Discrete Wavelet Packet Transform to extract the features related to different human vigilance states. We use well-known other cl...

2018 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE), 2018

Wireless Sensor Networks (WSNs) have become one of the most motivating research areas. In this pa... more Wireless Sensor Networks (WSNs) have become one of the most motivating research areas. In this paper we have proposed an Optimized Distributed Energy Efficient Clustering scheme for Heterogeneous WSNs (ODEEC) to prolong the network lifetime and stability period. This protocol proposes an optimization of the Cluster Head (CH) selection by modifying the probability function of heterogeneity. The network field is classified into two parts inner region and outer region. The sensor nodes decide to become cluster head nodes or not based on their positions from Base Station (BS) as well as the probabilities of heterogeneity model. Simulation results show that ODEEC achieves better performance among the relevant protocols in terms of throughput as well as network lifetime and stability period.

2017 International Conference on Communication, Control, Computing and Electronics Engineering (ICCCCEE), 2017

In this work an image processing system was developed to identify malaria parasites in thin blood... more In this work an image processing system was developed to identify malaria parasites in thin blood smears and to classify them into one of the four different species of malaria. Many techniques were implemented in the preprocessing stage to enhance the images. In the first part of the system morphological processing is applied to extract the Red Blood Cells (RBC) from blood images. The developed algorithm picks the suspicious regions and detects the parasites in the images including the overlapped cells. Accordingly, the RBCs are classified into infected and non-infected cells and the number of RBCs in each image is calculated. The second part of the system uses the Normalized Cross-Correlation function to classify the parasite into one of the four species namely, Plasmodium falciparum, Plasmodium vivax, Plasmodium ovale. Compared to manual results, the system achieved 95 % accuracy for detection and counting of RBCs and 100% for detection and classifying the parasite into one of its four types.

In this Thesis, parameters for a model of Arabic speech production were obtained. These parameter... more In this Thesis, parameters for a model of Arabic speech production were obtained. These parameters can serve in many applications such as speech recognition systems and voice response systems. The model parameters obtained are: 1. The prediction and reflection parameters, 2. The pitch period, 3. The gain factor, 4. The voice/unvoiced parameters.

2020 IEEE-EMBS Conference on Biomedical Engineering and Sciences (IECBES), 2021

Automating the process of human alertness classification has received special attention recently.... more Automating the process of human alertness classification has received special attention recently. Many applications can benefit from automatic alertness state identification, such as driver fatigue detection, monotonous task workers' vigilance detection and sleep studies in the medical field. Biological signals, such as Electrocardiogram (ECG), Electromyogram (EMG), and Electroencephalogram (EEG), along with human behaviors, such as head position, eye movement, have been used to infer the alertness state. Among all these, the EEG has been recognized as the best tool for this purpose. EEG is a complex signal that contains a wealth of information about brain’ activities. It is, however, vulnerable to noise and artifacts. Dimensionality reduction techniques, such as feature extraction and/or feature selection, can lead to a small number of robust features that can be used in many applications such automatic classifications. Entropy-based features have been widely used in EEG signal analysis. A new class of entropy measure, called Range Entropy (RangeEn), was recently proposed to address weaknesses of two widely used entropy measures, namely Approximate Entropy (ApEn) and Sample Entropy (SampEn). This paper aims at investigating the ability of RangeEn in discriminating between EEG behaviors associated with different human alertness states, namely awake, drowsy, and asleep.

The developing revolution in the technology of on-line storage, display, and communication will m... more The developing revolution in the technology of on-line storage, display, and communication will make it economically possible to place the entire contents of a library on-line, accessible from computer networks located any where, with a hardware cost comparable to operational budget of that library. A virtual engineering library is a collection of information that is stored and accessed electronically. The information stored in the library should have a number of topics common to all the data. At the same time, virtual engineering library represents a new infrastructure and interface that has been created by the integration and use of digital content. The objective of this research is to design and implement a virtual engineering library, and to examine the issues that effect and are affected by virtual engineering library development. The primary elements of this design are databases and interface tools. This framework provides a model of navigation that supports the process of discovery with predictable access points into differing collections of information. The implemented virtual engineering library works on Windows-Server operating systems and required Apache Web-server, PHP, and HTML to interact with the database (My SQL) of the proposed application. The research concludes by attempting to reach some conclusions about a virtual engineering library development and the validity of the proposed model of virtual engineering library. Flowing from this, recommendations are made for further research in this field.

2020 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE), 2021

This research aims to design and develop an accurate speech recognition system for a set of prede... more This research aims to design and develop an accurate speech recognition system for a set of predefined words collected from short audio clips. It uses The Speech Commands Dataset v0.01 provided by Google’s TensorFlow. Isolated word speech recognition can be implemented in voice user interfaces for applications with key-word spotting. The end goal is to classify and recognize ten words, along with classes for “unknown” words besides the “silence” class. The problems that face the current speech recognition technology like the acoustical noise and variations in recording environments are also solved and addressed here. To extract useful information from the signal, two methods of feature extraction were used: MFCCs and Mel-spectrograms. For classification, the convolutional neural network (CNN) was used. Different models were developed for this research, where each model has different architecture (1D-convnet and 2D-convnet). During training, techniques like batch normalization, regularization, and dropout were added to improve the accuracy and maintain the efficiency of the models. As a result of our experiments, The final model (2D-convnet with MFCC-16000) achieved an accuracy of 97.07% for training and 96.19% for testing.

مدختست ةرياعملا لمعم يف تارابتخلاا ةفرغ يف ةرارحلا ةجردو عضوملا يف مكحتلل ةرياعملا ةلواط ةدحو يون... more مدختست ةرياعملا لمعم يف تارابتخلاا ةفرغ يف ةرارحلا ةجردو عضوملا يف مكحتلل ةرياعملا ةلواط ةدحو يوناثلا (SSDL) . ةرياعملا ةلواط ةدحول يقيقحلا نمزلا يف ماظن ذيفنتو ميمصت وه لمعلا اذه نم فدهلا نإ كلا عضومل قيقد ديدحت ىلع لوصحلل جمدم مكحت ماظنب سايقلا ضرغب عاعشلإل ةضرعملا تانئا . ماظن لثميو ذفنملا ةرياعملا نوكتت ةيلآ ةدعاق نم حت ماظن ك و ةينورتكلا رئاودو قيقد م ةعومجم يكيناكيم ة . قيرط نع مكحتلا ي رابتخلاا ةفرغ جراخ مكحت ةحول نم ةرياعملا ةلواط يف ققح مامأ ةنيعلا عضومل بسانم ديدحت ماظنلا ةمزح اعشإ ايودي عضوملا ديدحتب ةنراقم ةيزآرم ع . ماظنلا جتني امآ ةفرغ ةرارح ةجردل ةبسانم تاءارق يعجرم ةرارح نازيم تاءارقب ةنراقم رابتخلاا . ي ؤ ليلقت و ةدوجلا ةيلاع ةرياعم تايلمع ىلإ ماظنلا اذه يد نمو عاعشلإل صاخشلأا ضرعت نمز نم ةياقولا لاجم يف ةديج ةمدخ ميدقت مث عاعشلإا . ABSTRACT The calibration Bench Module is used for controlling the position and monitoring the examination room temperature in Secondary Standard Dosimetry Labs (SSDLs). The objective of this work is to design and implement a real time calibration bench module with an embedded control system to provide a precise position for the objects to be irradiated for radiation metrology purpose. The implemented calibration bench represents a robotic platform, which consists of a microcontroller, electrical circuits and a mechanical assembly. By controlling the calibration bench from a control panel outside the examination room the system results suitable position allocation for the sample against the radiation central beam compared to manual positioning. It also results suitable readings of the examination room temperature compared to readings of a reference thermometer. The resultant system leads to high quality of the calibration process and less exposure time for the personnel i.e., good radiation protection.

2020 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE), 2021

This research aims to design and develop an accurate speech recognition system for a set of prede... more This research aims to design and develop an accurate speech recognition system for a set of predefined words collected from short audio clips. It uses The Speech Commands Dataset v0.01 provided by Google’s TensorFlow. Isolated word speech recognition can be implemented in voice user interfaces for applications with key-word spotting. The end
goal is to classify and recognize ten words, along with classes for
“unknown” words besides the “silence” class. The problems that
face the current speech recognition technology like the acoustical
noise and variations in recording environments are also solved
and addressed here. To extract useful information from the
signal, two methods of feature extraction were used: MFCCs and
Mel-spectrograms. For classification, the convolutional neural
network (CNN) was used. Different models were developed
for this research, where each model has a different architecture
(1D-convnet and 2D-convnet). During training, techniques like
batch normalization, regularization, and dropout were added to
improve the accuracy and maintain the efficiency of the models.
As a result of our experiments, The final model (2D-convnet
with MFCC-16000) achieved an accuracy of 97.07% for training
and 96.19% for testing.

This work was presented in COAR 2018 Annual Meeting and General Assembly, Hamburg (Germany)

Proceedings of the 5th WSEAS …, 2006

University of Khartoum Engineering Journal

2019 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE)

WIT Transactions on Information and Communication Technologies, 2002

WIT Transactions on Information and Communication Technologies, 1970

Journal of Biomedical Engineering and Medical Imaging, 2020

2018 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE), 2018

2017 International Conference on Communication, Control, Computing and Electronics Engineering (ICCCCEE), 2017

2020 IEEE-EMBS Conference on Biomedical Engineering and Sciences (IECBES), 2021

2020 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE), 2021

This research aims to design and develop an accurate speech recognition system for a set of prede... more This research aims to design and develop an accurate speech recognition system for a set of predefined words collected from short audio clips. It uses The Speech Commands Dataset v0.01 provided by Google’s TensorFlow. Isolated word speech recognition can be implemented in voice user interfaces for applications with key-word spotting. The end
goal is to classify and recognize ten words, along with classes for
“unknown” words besides the “silence” class. The problems that
face the current speech recognition technology like the acoustical
noise and variations in recording environments are also solved
and addressed here. To extract useful information from the
signal, two methods of feature extraction were used: MFCCs and
Mel-spectrograms. For classification, the convolutional neural
network (CNN) was used. Different models were developed
for this research, where each model has a different architecture
(1D-convnet and 2D-convnet). During training, techniques like
batch normalization, regularization, and dropout were added to
improve the accuracy and maintain the efficiency of the models.
As a result of our experiments, The final model (2D-convnet
with MFCC-16000) achieved an accuracy of 97.07% for training
and 96.19% for testing.

This work was presented in COAR 2018 Annual Meeting and General Assembly, Hamburg (Germany)

Proceedings of the 5th WSEAS …, 2006