Aya Ismail
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Aya Ismail has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Kamal Abdelraouf Eldahshan
Al-Azhar University
3 shared publications Marwa Elpeltagy
2 shared publications- Marwa Elpeltagy
Systems and Computers Department, Al-Azhar University, Cairo, Egypt
2 shared publications A. S. Salama
1 shared publicationAzaa Taha
Department of Mathematics, Faculty of Science Ain Shams University Cairo 11566 Egypt
1 shared publication- Doaa Aboshady
Mathematics Department, Tanta University, Tanta, Egypt
1 shared publication
Latest Publications
- Open Access
- 18 June 2025
- Published by Springer Nature in Neural Computing and Applications
- Vol. 37 (22) , 18527-18548
A novel lightweight dual-stream recurrent transformer model for anomaly detection in driving dashcam videos
In autonomous vehicles, it is significant to understand the surrounding environmental context to ensure reliability and safety. To achieve this, detecting a wide range of anomalies surrounding vehicles is crucial. This paper proposes a new robust and efficient model, called lightweight dual-stream recurrent transformer (LDSRT), for anomaly detection in driving dashcam videos. This model competes with state-of-the-art models, tackling their limited generalization problem, and providing reliable timely detection. The new model introduces an enhanced dual-stream network combining an improved MobileV3-Small network and an ameliorated convolution-3d network. This dual-stream design efficiently produces a compact representation of detailed static visual-spatial, temporal, and dynamic motion features. In addition, the model integrates a recurrent-enhanced transformer network for effectively understanding and learning the contextual relationships and temporal dependencies across driving video frames. This assures a comprehensive analysis of appearance and motion cues, both temporally and contextually. The model is then trained on a new integrated and diverse large dataset: DoTA-StreetAccident. The DoTA-StreetAccident dataset depends on merging two datasets: Detection of Traffic Anomaly (DoTA) and StreetAccident. Further, the LDSRT model’s effectiveness is extensively evaluated on the DoTA-StreetAccident and RetoTruck datasets, confirming its robustness across various driving anomaly detection tasks. Several evaluation metrics are utilized, and the analysis of experimental results and computational cost demonstrates that the proposed model significantly outperforms state-of-the-art models.
- 18 June 2025
- Published by Springer Nature in Neural Computing and Applications
- Vol. 37 (22) , 18527-18548
Open AccessA novel lightweight dual-stream recurrent transformer model for anomaly detection in driving dashcam videos
In autonomous vehicles, it is significant to understand the surrounding environmental context to ensure reliability and safety. To achieve this, detecting a wide range of anomalies surrounding vehicles is crucial. This paper proposes a new robust and efficient model, called lightweight dual-stream recurrent transformer (LDSRT), for anomaly detection in driving dashcam videos. This model competes with state-of-the-art models, tackling their limited generalization problem, and providing reliable timely detection. The new model introduces an enhanced dual-stream network combining an improved MobileV3-Small network and an ameliorated convolution-3d network. This dual-stream design efficiently produces a compact representation of detailed static visual-spatial, temporal, and dynamic motion features. In addition, the model integrates a recurrent-enhanced transformer network for effectively understanding and learning the contextual relationships and temporal dependencies across driving video frames. This assures a comprehensive analysis of appearance and motion cues, both temporally and contextually. The model is then trained on a new integrated and diverse large dataset: DoTA-StreetAccident. The DoTA-StreetAccident dataset depends on merging two datasets: Detection of Traffic Anomaly (DoTA) and StreetAccident. Further, the LDSRT model’s effectiveness is extensively evaluated on the DoTA-StreetAccident and RetoTruck datasets, confirming its robustness across various driving anomaly detection tasks. Several evaluation metrics are utilized, and the analysis of experimental results and computational cost demonstrates that the proposed model significantly outperforms state-of-the-art models.
- Open Access
- 26 January 2024
- Published by Qeios Ltd
Review of: "Comparing YOLOv8 and Mask RCNN for object segmentation in complex orchard environments"
THE PAPER, ENTITLED “COMPARING YOLOV8 AND MASK RCNN FOR OBJECT SEGMENTATION IN COMPLEX ORCHARD ENVIRONMENTS”, COMPARES THE ONE-STAGE YOLOV8 AND THE TWO-STAGE MASK R-CNN MACHINE LEARNING MODELS FOR INSTANCE SEGMENTATION UNDER VARYING ORCHARD CONDITION...
- 26 January 2024
- Published by Qeios Ltd
Open AccessReview of: "Comparing YOLOv8 and Mask RCNN for object segmentation in complex orchard environments"
THE PAPER, ENTITLED “COMPARING YOLOV8 AND MASK RCNN FOR OBJECT SEGMENTATION IN COMPLEX ORCHARD ENVIRONMENTS”, COMPARES THE ONE-STAGE YOLOV8 AND THE TWO-STAGE MASK R-CNN MACHINE LEARNING MODELS FOR INSTANCE SEGMENTATION UNDER VARYING ORCHARD CONDITION...
- Open Access
- 2 August 2022
- Published by Springer Nature in Neural Computing and Applications
- Vol. 34 (24) , 21777-21791
An integrated spatiotemporal-based methodology for deepfake detection
Rapid advances in deep learning models have made it easier for public and crackers to generate hyper-realistic deepfake videos in which faces are swapped. Such deepfake videos may constitute a significant threat to the world if they are misused to blackmail public figures and to deceive systems of face recognition. As a result, distinguishing these fake videos from real ones has become fundamental. This paper introduces a new deepfake video detection method. You Only Look Once (YOLO) face detector is used to detect faces from video frames. A proposed hybrid method based on proposing two different feature extraction methods is applied to these faces. The first feature extraction method, a proposed Convolution Neural Network (CNN), is based on the Histogram of Oriented Gradient (HOG) method. The second one is an ameliorated XceptionNet CNN. The two extracted sets of features are merged together and fed as input to a sequence of Gated Recurrent Units (GRUs) to extract the spatial and temporal features and then individuate the authenticity of videos. The proposed method is trained on the CelebDF-FaceForencics++ (c23) dataset and evaluated on the CelebDF test set. The experimental results and analysis confirm the superiority of the suggested method over the state-of-the-art methods.
- 2 August 2022
- Published by Springer Nature in Neural Computing and Applications
- Vol. 34 (24) , 21777-21791
Open AccessAn integrated spatiotemporal-based methodology for deepfake detection
Rapid advances in deep learning models have made it easier for public and crackers to generate hyper-realistic deepfake videos in which faces are swapped. Such deepfake videos may constitute a significant threat to the world if they are misused to blackmail public figures and to deceive systems of face recognition. As a result, distinguishing these fake videos from real ones has become fundamental. This paper introduces a new deepfake video detection method. You Only Look Once (YOLO) face detector is used to detect faces from video frames. A proposed hybrid method based on proposing two different feature extraction methods is applied to these faces. The first feature extraction method, a proposed Convolution Neural Network (CNN), is based on the Histogram of Oriented Gradient (HOG) method. The second one is an ameliorated XceptionNet CNN. The two extracted sets of features are merged together and fed as input to a sequence of Gated Recurrent Units (GRUs) to extract the spatial and temporal features and then individuate the authenticity of videos. The proposed method is trained on the CelebDF-FaceForencics++ (c23) dataset and evaluated on the CelebDF test set. The experimental results and analysis confirm the superiority of the suggested method over the state-of-the-art methods.
- Open Access
- 21 September 2021
- Published by PeerJ in PeerJ Computer Science
- Vol. 7, e730
Deepfake video detection: YOLO-Face convolution recurrent approach
Recently, the deepfake techniques for swapping faces have been spreading, allowing easy creation of hyper-realistic fake videos. Detecting the authenticity of a video has become increasingly critical because of the potential negative impact on the world. Here, a new project is introduced; You Only Look Once Convolution Recurrent Neural Networks (YOLO-CRNNs), to detect deepfake videos. The YOLO-Face detector detects face regions from each frame in the video, whereas a fine-tuned EfficientNet-B5 is used to extract the spatial features of these faces. These features are fed as a batch of input sequences into a Bidirectional Long Short-Term Memory (Bi-LSTM), to extract the temporal features. The new scheme is then evaluated on a new large-scale dataset; CelebDF-FaceForencics++ (c23), based on a combination of two popular datasets; FaceForencies++ (c23) and Celeb-DF. It achieves an Area Under the Receiver Operating Characteristic Curve (AUROC) 89.35% score, 89.38% accuracy, 83.15% recall, 85.55% precision, and 84.33% F1-measure for pasting data approach. The experimental analysis approves the superiority of the proposed method compared to the state-of-the-art methods.
- 21 September 2021
- Published by PeerJ in PeerJ Computer Science
- Vol. 7, e730
Open AccessDeepfake video detection: YOLO-Face convolution recurrent approach
Recently, the deepfake techniques for swapping faces have been spreading, allowing easy creation of hyper-realistic fake videos. Detecting the authenticity of a video has become increasingly critical because of the potential negative impact on the world. Here, a new project is introduced; You Only Look Once Convolution Recurrent Neural Networks (YOLO-CRNNs), to detect deepfake videos. The YOLO-Face detector detects face regions from each frame in the video, whereas a fine-tuned EfficientNet-B5 is used to extract the spatial features of these faces. These features are fed as a batch of input sequences into a Bidirectional Long Short-Term Memory (Bi-LSTM), to extract the temporal features. The new scheme is then evaluated on a new large-scale dataset; CelebDF-FaceForencics++ (c23), based on a combination of two popular datasets; FaceForencies++ (c23) and Celeb-DF. It achieves an Area Under the Receiver Operating Characteristic Curve (AUROC) 89.35% score, 89.38% accuracy, 83.15% recall, 85.55% precision, and 84.33% F1-measure for pasting data approach. The experimental analysis approves the superiority of the proposed method compared to the state-of-the-art methods.
- Open Access
A New Deep Learning-Based Methodology for Video Deepfake Detection Using XGBoost
Currently, face-swapping deepfake techniques are widely spread, generating a significant number of highly realistic fake videos that threaten the privacy of people and countries. Due to their devastating impacts on the world, distinguishing between real and deepfake videos has become a fundamental issue. This paper presents a new deepfake detection method: you only look once–convolutional neural network–extreme gradient boosting (YOLO-CNN-XGBoost). The YOLO face detector is employed to extract the face area from video frames, while the InceptionResNetV2 CNN is utilized to extract features from these faces. These features are fed into the XGBoost that works as a recognizer on the top level of the CNN network. The proposed method achieves 90.62% of an area under the receiver operating characteristic curve (AUC), 90.73% accuracy, 93.53% specificity, 85.39% sensitivity, 85.39% recall, 87.36% precision, and 86.36% F1-measure on the CelebDF-FaceForencics++ (c23) merged dataset. The experimental study confirms the superiority of the presented method as compared to the state-of-the-art methods.
Open AccessA New Deep Learning-Based Methodology for Video Deepfake Detection Using XGBoost
Currently, face-swapping deepfake techniques are widely spread, generating a significant number of highly realistic fake videos that threaten the privacy of people and countries. Due to their devastating impacts on the world, distinguishing between real and deepfake videos has become a fundamental issue. This paper presents a new deepfake detection method: you only look once–convolutional neural network–extreme gradient boosting (YOLO-CNN-XGBoost). The YOLO face detector is employed to extract the face area from video frames, while the InceptionResNetV2 CNN is utilized to extract features from these faces. These features are fed into the XGBoost that works as a recognizer on the top level of the CNN network. The proposed method achieves 90.62% of an area under the receiver operating characteristic curve (AUC), 90.73% accuracy, 93.53% specificity, 85.39% sensitivity, 85.39% recall, 87.36% precision, and 86.36% F1-measure on the CelebDF-FaceForencics++ (c23) merged dataset. The experimental study confirms the superiority of the presented method as compared to the state-of-the-art methods.
- 14 December 2017
- Published by Wiley in Mathematical Methods in the Applied Sciences
- Vol. 41 (15) , 5780-5791
A new retrieval method based on time series variation using field association terms
This paper studies the effects of the time change on the frequencies of specific terms connected to the document field in a given period. These specific terms are the field association (FA) terms. The paper also suggests a new method for automatic evaluation of the stabilization classes of FA terms to improve the precision of decision tree. The stabilization classes point out the popularity of list of FA terms depending on time change. Moreover, the suggested method manipulates the problem of the scattering of data numbers among classes to improve the performance of decision tree precision. The presented method is evaluated through conducting experiments by simulating the result of 1245 files, which are equivalent to 4.15 MB. The F‐measure for increment, fairly steady, and decrement classes achieves %90.4, %99.3, and %38.6, sequentially.
- 14 December 2017
- Published by Wiley in Mathematical Methods in the Applied Sciences
- Vol. 41 (15) , 5780-5791
A new retrieval method based on time series variation using field association terms
This paper studies the effects of the time change on the frequencies of specific terms connected to the document field in a given period. These specific terms are the field association (FA) terms. The paper also suggests a new method for automatic evaluation of the stabilization classes of FA terms to improve the precision of decision tree. The stabilization classes point out the popularity of list of FA terms depending on time change. Moreover, the suggested method manipulates the problem of the scattering of data numbers among classes to improve the performance of decision tree precision. The presented method is evaluated through conducting experiments by simulating the result of 1245 files, which are equivalent to 4.15 MB. The F‐measure for increment, fairly steady, and decrement classes achieves %90.4, %99.3, and %38.6, sequentially.