Preferred Language
Articles
/
alkej-814
Comparative Transfer Learning Models for End-to-End Self-Driving Car
...Show More Authors

Self-driving automobiles are prominent in science and technology, which affect social and economic development. Deep learning (DL) is the most common area of study in artificial intelligence (AI). In recent years, deep learning-based solutions have been presented in the field of self-driving cars and have achieved outstanding results. Different studies investigated a variety of significant technologies for autonomous vehicles, including car navigation systems, path planning, environmental perception, as well as car control. End-to-end learning control directly converts sensory data into control commands in autonomous driving. This research aims to identify the most accurate pre-trained Deep Neural Network (DNN) for predicting the steering angle of a self-driving vehicle that is suitable to be applied to embedded automotive technologies with limited performance. Three well-known pre-trained models were compared in this study: AlexNet, ResNet18, and DenseNet121.

Transfer learning was utilized by modifying the final layer of pre-trained models in order to predict the steering angle of the vehicle. Experiments were conducted on the dataset collected from two different tracks. According to the study's findings, ResNet18 and DenseNet121 have the lowest error percentage for steering angle values. Furthermore, the performance of the modified models was evaluated on predetermined tracks. ResNet18 outperformed DenseNet121 in terms of accuracy, with less deviation from the center of the track, while DenseNet121 demonstrated greater adaptability across multiple tracks, resulting in better performance consistency.

Scopus Crossref
View Publication Preview PDF
Quick Preview PDF
Publication Date
Mon Dec 05 2022
Journal Name
Baghdad Science Journal
Normal Self-injective Hyperrings
...Show More Authors

       In this paper normal self-injective hyperrings are introduced and studied. Some new relations between this concept and essential hyperideal, dense hyperideal, and divisible hyperring are studied. 

View Publication Preview PDF
Scopus Crossref
Publication Date
Sat Jan 19 2019
Journal Name
Artificial Intelligence Review
Survey on supervised machine learning techniques for automatic text classification
...Show More Authors

View Publication
Scopus (300)
Crossref (271)
Scopus Clarivate Crossref
Publication Date
Sun Jan 01 2023
Journal Name
Computers, Materials & Continua
Hybrid Deep Learning Enabled Load Prediction for Energy Storage Systems
...Show More Authors

View Publication
Scopus (16)
Crossref (13)
Scopus Clarivate Crossref
Publication Date
Thu Mar 13 2025
Journal Name
Academia Open
Deep Learning and Fusion Techniques for High-Precision Image Matting:
...Show More Authors

General Background: Deep image matting is a fundamental task in computer vision, enabling precise foreground extraction from complex backgrounds, with applications in augmented reality, computer graphics, and video processing. Specific Background: Despite advancements in deep learning-based methods, preserving fine details such as hair and transparency remains a challenge. Knowledge Gap: Existing approaches struggle with accuracy and efficiency, necessitating novel techniques to enhance matting precision. Aims: This study integrates deep learning with fusion techniques to improve alpha matte estimation, proposing a lightweight U-Net model incorporating color-space fusion and preprocessing. Results: Experiments using the AdobeComposition-1k

... Show More
View Publication Preview PDF
Publication Date
Thu Jun 01 2023
Journal Name
International Journal Of Electrical And Computer Engineering (ijece)
An optimized deep learning model for optical character recognition applications
...Show More Authors

The convolutional neural networks (CNN) are among the most utilized neural networks in various applications, including deep learning. In recent years, the continuing extension of CNN into increasingly complicated domains has made its training process more difficult. Thus, researchers adopted optimized hybrid algorithms to address this problem. In this work, a novel chaotic black hole algorithm-based approach was created for the training of CNN to optimize its performance via avoidance of entrapment in the local minima. The logistic chaotic map was used to initialize the population instead of using the uniform distribution. The proposed training algorithm was developed based on a specific benchmark problem for optical character recog

... Show More
View Publication
Scopus (1)
Scopus Crossref
Publication Date
Tue Apr 30 2024
Journal Name
International Journal On Technical And Physical Problems Of Engineering
Deep Learning Techniques For Skull Stripping of Brain MR Images
...Show More Authors

Deep Learning Techniques For Skull Stripping of Brain MR Images

Scopus (1)
Scopus
Publication Date
Mon Dec 18 2017
Journal Name
Al-khwarizmi Engineering Journal
Comparative Study of Performance and Emission Characteristics between Spark Ignition Engine and Homogeneous Charge Compression Ignition Engine (HCCI)
...Show More Authors

Many researchers consider Homogeneous Charge Compression Ignition (HCCI) engine mode as a promising alternative to combustion in Spark Ignition and Compression Ignition Engines. The HCCI engine runs on lean mixtures of fuel and air, and the combustion is produced from the fuel autoignition instead of ignited by a spark. This combustion mode was investigated in this paper. A variable compression ratio, spark ignition engine type TD110 was used in the experiments. The tested fuel was Iraqi conventional gasoline (ON=82).

The results showed that HCCI engine can run in very lean equivalence ratios. The brake specific fuel consumption was reduced about 28% compared with a spark ignition engine. The experimental tests showed that the em

... Show More
View Publication Preview PDF
Crossref
Publication Date
Wed Jul 20 2022
Journal Name
Ibn Al-haitham Journal For Pure And Applied Sciences
A Theoretical Investigation of Charge Transfer Dynamics from Sensitized Molecule D35CPDT Dye to SnO_2 and TiO_2 Semiconductor
...Show More Authors

In this research, the dynamics process of charge transfer from the sensitized  D35CPDT dye to tin(iv) oxide( ) or titanium dioxide (  ) semiconductors are carried out by using a quantum model for charge transfer. Different chemical solvents Pyridine, 2-Methoxyethanol. Ethanol, Acetonitrile, and Methanol have been used with both systems as polar media surrounded the systems. The rate for charge transfer from photo-excitation D35CPDTdye and injection into the conduction band of  or  semiconductors vary from a  to  for system and from a   to  for the system, depending on the charge transfer parameters strength coupling, free energy, potential of donor and acceptor in the system. The charge transfer rate in D35CPDT /  the syst

... Show More
View Publication
Crossref
Publication Date
Tue Dec 01 2015
Journal Name
Journal Of Engineering
Modified Grid Clustering Technique to Predict Heat Transfer Coefficient in a Duct of Arbitrary Cross Section Area
...Show More Authors

A simple straightforward mathematical method has been developed to cluster grid nodes on a boundary segment of an arbitrary geometry that can be fitted by a relevant polynomial. The method of solution is accomplished in two steps. At the first step, the length of the boundary segment is evaluated by using the mean value theorem, then grids are clustered as desired, using relevant linear clustering functions. At the second step, as the coordinates cell nodes have been computed and the incremental distance between each two nodes has been evaluated, the original coordinate of each node is then computed utilizing the same fitted polynomial with the mean value theorem but reversibly.

The method is utilized to predict

... Show More
View Publication Preview PDF
Publication Date
Sat Aug 01 2020
Journal Name
Periodicals Of Engineering And Natural Sciences
A comparison of some forecasting models to forecast the number of old people in Iraqi retirement homes
...Show More Authors

Statistical methods of forecasting have applied with the intention of constructing a model to predict the number of the old aged people in retirement homes in Iraq. They were based on the monthly data of old aged people in Baghdad and the governorates except for the Kurdistan region from 2016 to 2019. Using BoxJenkins methodology, the stationarity of the series was examined. The appropriate model order was determined, the parameters were estimated, the significance was tested, adequacy of the model was checked, and then the best model of prediction was used. The best model for forecasting according to criteria of (Normalized BIC, MAPE, RMSE) is ARIMA (0, 1, 2)