Moamen Zaher
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Moamen Zaher has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Amr S. Ghoneim
Helwan University
4 shared publications - Ayman Atia
Department of Computer Science, October University for Modern Sciences and Arts (MSA), Giza, Egypt
4 shared publications - Laila Abdelhamid
Information System Dept. Faculty of Computing and Artificial Intelligence, Helwan University, Cairo, Egypt
4 shared publications Ammar Mohamed
October University for Modern Sciences and Arts
2 shared publications- Ahmed Diaaeldin
October University for Modern Sciences and Arts (MSA),Faculty of Computer Science,Giza,Egypt
1 shared publication Ahmed El-Darawi
October University for Modern Sciences and Arts
1 shared publication
Latest Publications
- 24 March 2026
- Published by Springer Nature
- p. 531-536
Affective Recommendation Through Physiological and Multimodal User Modeling
- 24 March 2026
- Published by Springer Nature
- p. 531-536
Affective Recommendation Through Physiological and Multimodal User Modeling
- 17 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 489-496
Improving Tele-Rehabilitation Performance: The Role of Ensemble Learning in Human Activity Recognition
Physical rehabilitation is essential for restoring mobility, reducing pain, and improving quality of life after injury or surgical interventions. Accurate Human Activity Recognition (HAR) is critical in this context, yet existing models often struggle to generalize across diverse rehabilitation tasks and patient populations. Therefore, we first developed a set of standalone models including: Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN) and a hybrid CNN-LSTM to establish baseline performance on sensor-based (UCI-HAR) and camera-based skeletal (UI-PRMD) datasets. Building on these baselines, we investigated ensemble learning techniques-including majority voting, averaging, and stacking-to enhance robustness and classification accuracy. The results demonstrate clear performance gains: our approach achieved 99.16 % accuracy on UCI-HAR using a CNN-based meta-learner, and 99.75 % on UI-PRMD with a weighted ensemble strategy. These findings underscore the potential of ensemble methods to outperform individual models, offering a pathway toward more reliable rehabilitation monitoring, realtime feedback, and scalable remote therapy.
- 17 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 489-496
Improving Tele-Rehabilitation Performance: The Role of Ensemble Learning in Human Activity Recognition
Physical rehabilitation is essential for restoring mobility, reducing pain, and improving quality of life after injury or surgical interventions. Accurate Human Activity Recognition (HAR) is critical in this context, yet existing models often struggle to generalize across diverse rehabilitation tasks and patient populations. Therefore, we first developed a set of standalone models including: Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN) and a hybrid CNN-LSTM to establish baseline performance on sensor-based (UCI-HAR) and camera-based skeletal (UI-PRMD) datasets. Building on these baselines, we investigated ensemble learning techniques-including majority voting, averaging, and stacking-to enhance robustness and classification accuracy. The results demonstrate clear performance gains: our approach achieved 99.16 % accuracy on UCI-HAR using a CNN-based meta-learner, and 99.75 % on UI-PRMD with a weighted ensemble strategy. These findings underscore the potential of ensemble methods to outperform individual models, offering a pathway toward more reliable rehabilitation monitoring, realtime feedback, and scalable remote therapy.
- Open Access
- 4 February 2025
- Published by Taylor & Francis in Journal of Information and Telecommunication
- Vol. 9 (3) , 382-402
Rehabilitation monitoring and assessment: a comparative analysis of feature engineering and machine learning algorithms on the UI-PRMD and KIMORE benchmark datasets
- 4 February 2025
- Published by Taylor & Francis in Journal of Information and Telecommunication
- Vol. 9 (3) , 382-402
Open AccessRehabilitation monitoring and assessment: a comparative analysis of feature engineering and machine learning algorithms on the UI-PRMD and KIMORE benchmark datasets
- 27 November 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 184, 109399
Fusing CNNs and attention-mechanisms to improve real-time indoor Human Activity Recognition for classifying home-based physical rehabilitation exercises
- 27 November 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 184, 109399
Fusing CNNs and attention-mechanisms to improve real-time indoor Human Activity Recognition for classifying home-based physical rehabilitation exercises
- 13 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 336-341
Referring Video Object Clustering (RVOC)
System surveillance involves the continuous monitoring and analysis of organizational data to ensure security and operational efficiency. This proactive approach aids in detecting anomalies, threats, and performance issues, thereby safeguarding sensitive information and maintaining system integrity. This paper proposes a framework for querying videos named Referring Video Object Clustering (RVOC). Traditional methods necessitate identifying the number of clusters before object identification based on pre-trained datasets, with components such as Principal Component Analysis (PCA) relying on fixed numerical values. In contrast, The proposed framework (RVOC) performs dynamic clustering without prior training, allowing the number of classes to adjust based on retrieved video objects. This method employs a sophisticated NLP query system, enabling intricate searches like “person skating on a red skateboard.” The system diligently searches and analyzes video content to find instances matching the query, grouping and segmenting detected individuals with advanced clustering algorithms. This enhances user experience by facilitating easy navigation and selection of desired individuals, with each cluster representing a unique person or object. RVOC's efficacy is evidenced by its Normalized Mutual Information (NMI) score of 51% in tests with multilabel car and color datasets.
- 13 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 336-341
Referring Video Object Clustering (RVOC)
System surveillance involves the continuous monitoring and analysis of organizational data to ensure security and operational efficiency. This proactive approach aids in detecting anomalies, threats, and performance issues, thereby safeguarding sensitive information and maintaining system integrity. This paper proposes a framework for querying videos named Referring Video Object Clustering (RVOC). Traditional methods necessitate identifying the number of clusters before object identification based on pre-trained datasets, with components such as Principal Component Analysis (PCA) relying on fixed numerical values. In contrast, The proposed framework (RVOC) performs dynamic clustering without prior training, allowing the number of classes to adjust based on retrieved video objects. This method employs a sophisticated NLP query system, enabling intricate searches like “person skating on a red skateboard.” The system diligently searches and analyzes video content to find instances matching the query, grouping and segmenting detected individuals with advanced clustering algorithms. This enhances user experience by facilitating easy navigation and selection of desired individuals, with each cluster representing a unique person or object. RVOC's efficacy is evidenced by its Normalized Mutual Information (NMI) score of 51% in tests with multilabel car and color datasets.
- 13 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 416-422
Enhancing Road Safety: Leveraging CNN-LSTM and Bi-LSTM Models for Advanced Driver Behavior Detection
Recognizing the importance of driver behavior is essential for enhancing road safety and optimizing traffic management systems. This study employs advanced deep learning techniques, specifically CNN-LSTM and Bi-LSTM models, to refine the prediction of driver behaviors using sensor data from the Honda Research Institute Driving Dataset (HDD). Our approach integrates a robust dataset encompassing a broad spectrum of sensor inputs, from vehicle dynamics to driver operational parameters, propelling advancements in driver behavior detection. The methodologies utilized enable the discernment of subtle and complex driving patterns, contributing to the reduction of road safety hazards. Our findings indicate that these models significantly improve the detection of hazardous driving behaviors, surpassing previous state-of-the-art methodologies with notable gains in mean average precision (mAP). These advancements affirm the potential of deep learning technologies in crafting sophisticated predictive safety systems, paving the way for future innovations.
- 13 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 416-422
Enhancing Road Safety: Leveraging CNN-LSTM and Bi-LSTM Models for Advanced Driver Behavior Detection
Recognizing the importance of driver behavior is essential for enhancing road safety and optimizing traffic management systems. This study employs advanced deep learning techniques, specifically CNN-LSTM and Bi-LSTM models, to refine the prediction of driver behaviors using sensor data from the Honda Research Institute Driving Dataset (HDD). Our approach integrates a robust dataset encompassing a broad spectrum of sensor inputs, from vehicle dynamics to driver operational parameters, propelling advancements in driver behavior detection. The methodologies utilized enable the discernment of subtle and complex driving patterns, contributing to the reduction of road safety hazards. Our findings indicate that these models significantly improve the detection of hazardous driving behaviors, surpassing previous state-of-the-art methodologies with notable gains in mean average precision (mAP). These advancements affirm the potential of deep learning technologies in crafting sophisticated predictive safety systems, paving the way for future innovations.
- Open Access
- 12 April 2024
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (3) , 1261-1301
Unlocking the potential of RNN and CNN models for accurate rehabilitation exercise classification on multi-datasets
Physical rehabilitation is crucial in healthcare, facilitating recovery from injuries or illnesses and improving overall health. However, a notable global challenge stems from the shortage of professional physiotherapists, particularly acute in some developing countries, where the ratio can be as low as one physiotherapist per 100,000 individuals. To address these challenges and elevate patient care, the field of physical rehabilitation is progressively integrating Computer Vision and Human Activity Recognition (HAR) techniques. Numerous research efforts aim to explore methodologies that assist in rehabilitation exercises and evaluate patient movements, which is crucial as incorrect exercises can potentially worsen conditions. This study investigates applying various deep-learning models for classifying exercises using the benchmark KIMORE and UI-PRMD datasets. Employing Bi-LSTM, LSTM, CNN, and CNN-LSTM, alongside a Random Search for architectural design and Hyper-parameter tuning, our investigation reveals the (CNN) model as the top performer. After applying cross-validation, the technique achieves remarkable mean testing accuracy rates of 93.08% on the KIMORE dataset and 99.7% on the UI-PRMD dataset. This marks a slight improvement of 0.75% and 0.1%, respectively, compared to previous techniques. In addition, expanding beyond exercise classification, this study explores the KIMORE dataset’s utility for disease identification, where the (CNN) model consistently demonstrates an outstanding accuracy of 89.87%, indicating its promising role in both exercises and disease identification within the context of physical rehabilitation.
- 12 April 2024
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 84 (3) , 1261-1301
Open AccessUnlocking the potential of RNN and CNN models for accurate rehabilitation exercise classification on multi-datasets
Physical rehabilitation is crucial in healthcare, facilitating recovery from injuries or illnesses and improving overall health. However, a notable global challenge stems from the shortage of professional physiotherapists, particularly acute in some developing countries, where the ratio can be as low as one physiotherapist per 100,000 individuals. To address these challenges and elevate patient care, the field of physical rehabilitation is progressively integrating Computer Vision and Human Activity Recognition (HAR) techniques. Numerous research efforts aim to explore methodologies that assist in rehabilitation exercises and evaluate patient movements, which is crucial as incorrect exercises can potentially worsen conditions. This study investigates applying various deep-learning models for classifying exercises using the benchmark KIMORE and UI-PRMD datasets. Employing Bi-LSTM, LSTM, CNN, and CNN-LSTM, alongside a Random Search for architectural design and Hyper-parameter tuning, our investigation reveals the (CNN) model as the top performer. After applying cross-validation, the technique achieves remarkable mean testing accuracy rates of 93.08% on the KIMORE dataset and 99.7% on the UI-PRMD dataset. This marks a slight improvement of 0.75% and 0.1%, respectively, compared to previous techniques. In addition, expanding beyond exercise classification, this study explores the KIMORE dataset’s utility for disease identification, where the (CNN) model consistently demonstrates an outstanding accuracy of 89.87%, indicating its promising role in both exercises and disease identification within the context of physical rehabilitation.