Hoda Mohammadzade
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Hoda Mohammadzade has a total of 39 co-authors. Here are the top co-authors by number of shared publications:
Benyamin Ghojogh
5 shared publicationsMohsen Tabejamaat
Electrical Engineering Department, Sharif University of Technology, Tehran, Iran 11155-8639
4 shared publications- Hamid Behroozi
Electrical Engineering Department, Sharif University of Technology, Tehran, Iran
3 shared publications - Alireza Nourbakhsh
Department of Electrical Engineering, Sharif University of Technology, Tehran, Iran
2 shared publications - Benyamin Ghojogh2 shared publications
- Hassan Keshvarikhojasteh
Sharif University of Technology Tehran Iran
2 shared publications
Latest Publications
- 15 June 2026
- Published by Elsevier in Journal of Visual Communication and Image Representation
- Vol. 119
A low-computational video synopsis framework and a benchmark dataset
- 15 June 2026
- Published by Elsevier in Journal of Visual Communication and Image Representation
- Vol. 119
A low-computational video synopsis framework and a benchmark dataset
- Open Access
Human-Centric Anomaly Detection in Surveillance Videos Using YOLO-World and Spatio-Temporal Deep Learning
Anomaly detection in surveillance videos remains a challenging task due to the diversity of abnormal events, class imbalance, and scene-dependent visual clutter. To address these issues, we propose a robust deep learning framework that integrates human-centric preprocessing with spatio-temporal modeling for multi-class anomaly classification. Our pipeline begins by applying YOLO-World - an open-vocabulary vision-language detector - to identify human instances in raw video clips, followed by ByteTrack for consistent identity-aware tracking. Background regions outside detected bounding boxes are suppressed via Gaussian blurring, effectively reducing scene-specific distractions and focusing the model on behaviorally relevant foreground content. The refined frames are then processed by an ImageNet-pretrained InceptionV3 network for spatial feature extraction, and temporal dynamics are captured using a bidirectional LSTM (BiLSTM) for sequence-level classification. Evaluated on a five-class subset of the UCF-Crime dataset (Normal, Burglary, Fighting, Arson, Explosion), our method achieves a mean test accuracy of 92.41% across three independent trials, with per-class F1-scores consistently exceeding 0.85. Comprehensive evaluation metrics - including confusion matrices, ROC curves, and macro/weighted averages - demonstrate strong generalization and resilience to class imbalance. The results confirm that foreground-focused preprocessing significantly enhances anomaly discrimination in real-world surveillance scenarios.
Open AccessHuman-Centric Anomaly Detection in Surveillance Videos Using YOLO-World and Spatio-Temporal Deep Learning
Anomaly detection in surveillance videos remains a challenging task due to the diversity of abnormal events, class imbalance, and scene-dependent visual clutter. To address these issues, we propose a robust deep learning framework that integrates human-centric preprocessing with spatio-temporal modeling for multi-class anomaly classification. Our pipeline begins by applying YOLO-World - an open-vocabulary vision-language detector - to identify human instances in raw video clips, followed by ByteTrack for consistent identity-aware tracking. Background regions outside detected bounding boxes are suppressed via Gaussian blurring, effectively reducing scene-specific distractions and focusing the model on behaviorally relevant foreground content. The refined frames are then processed by an ImageNet-pretrained InceptionV3 network for spatial feature extraction, and temporal dynamics are captured using a bidirectional LSTM (BiLSTM) for sequence-level classification. Evaluated on a five-class subset of the UCF-Crime dataset (Normal, Burglary, Fighting, Arson, Explosion), our method achieves a mean test accuracy of 92.41% across three independent trials, with per-class F1-scores consistently exceeding 0.85. Comprehensive evaluation metrics - including confusion matrices, ROC curves, and macro/weighted averages - demonstrate strong generalization and resilience to class imbalance. The results confirm that foreground-focused preprocessing significantly enhances anomaly discrimination in real-world surveillance scenarios.
- Open Access
- 5 August 2025
- Published by SCI AND TECH UNIVERSAL INC in Scientia Iranica
Deep Time Warping for Multiple Time Series Alignment
Time Series Alignment is a crucial task in signal processing with wide-ranging applications. Real-world signals often suffer from temporal shifts and scaling, leading to errors in raw data classification. This paper presents a novel Deep Learning-based approach for Multiple Time Series Alignment (MTSA). While existing methods mainly focus on Multiple Sequence Alignment (MSA) for biological sequences, there is a notable lack of alignment techniques for numerical time series. Traditional methods also typically address pairwise alignment, whereas our approach aligns all signals simultaneously, improving both alignment efficiency and computational speed. By decomposing to piece-wise linear sections, we introduce varying complexity into the warping function while ensuring compliance with three key constraints: boundary, monotonicity, and continuity conditions. We propose a deep convolutional network with a novel loss function that addresses key limitations of Dynamic Time Warping (DTW). Experiments on the UCR Archive 2018, involving 129 time series datasets, show that our method significantly enhances classification accuracy, warping average, and runtime efficiency across most datasets.
- 5 August 2025
- Published by SCI AND TECH UNIVERSAL INC in Scientia Iranica
Open AccessDeep Time Warping for Multiple Time Series Alignment
Time Series Alignment is a crucial task in signal processing with wide-ranging applications. Real-world signals often suffer from temporal shifts and scaling, leading to errors in raw data classification. This paper presents a novel Deep Learning-based approach for Multiple Time Series Alignment (MTSA). While existing methods mainly focus on Multiple Sequence Alignment (MSA) for biological sequences, there is a notable lack of alignment techniques for numerical time series. Traditional methods also typically address pairwise alignment, whereas our approach aligns all signals simultaneously, improving both alignment efficiency and computational speed. By decomposing to piece-wise linear sections, we introduce varying complexity into the warping function while ensuring compliance with three key constraints: boundary, monotonicity, and continuity conditions. We propose a deep convolutional network with a novel loss function that addresses key limitations of Dynamic Time Warping (DTW). Experiments on the UCR Archive 2018, involving 129 time series datasets, show that our method significantly enhances classification accuracy, warping average, and runtime efficiency across most datasets.
- Open Access
Deep Time Warping for Multiple Time Series Alignment
Time Series Alignment is a critical task in signal processing with numerous real-world applications. In practice, signals often exhibit temporal shifts and scaling, making classification on raw data prone to errors. This paper introduces a novel approach for Multiple Time Series Alignment (MTSA) leveraging Deep Learning techniques. While most existing methods primarily address Multiple Sequence Alignment (MSA) for protein and DNA sequences, there remains a significant gap in alignment methodologies for numerical time series. Additionally, conventional approaches typically focus on pairwise alignment, whereas our proposed method aligns all signals in a multiple manner (all the signals are aligned together at once). This innovation not only enhances alignment efficiency but also significantly improves computational speed. By decomposing into piece-wise linear sections, we introduce varying levels of complexity into the warping function. Additionally, our method ensures the satisfaction of three warping constraints: boundary, monotonicity, and continuity conditions. The utilization of a deep convolutional network allows us to employ a new loss function, addressing some limitations of Dynamic Time Warping (DTW). Experimental results on the UCR Archive 2018, comprising 129 time series datasets, demonstrate that employing our approach to align signals significantly enhances classification accuracy and warping average and also reduces the run time across the majority of these datasets.
Open AccessDeep Time Warping for Multiple Time Series Alignment
Time Series Alignment is a critical task in signal processing with numerous real-world applications. In practice, signals often exhibit temporal shifts and scaling, making classification on raw data prone to errors. This paper introduces a novel approach for Multiple Time Series Alignment (MTSA) leveraging Deep Learning techniques. While most existing methods primarily address Multiple Sequence Alignment (MSA) for protein and DNA sequences, there remains a significant gap in alignment methodologies for numerical time series. Additionally, conventional approaches typically focus on pairwise alignment, whereas our proposed method aligns all signals in a multiple manner (all the signals are aligned together at once). This innovation not only enhances alignment efficiency but also significantly improves computational speed. By decomposing into piece-wise linear sections, we introduce varying levels of complexity into the warping function. Additionally, our method ensures the satisfaction of three warping constraints: boundary, monotonicity, and continuity conditions. The utilization of a deep convolutional network allows us to employ a new loss function, addressing some limitations of Dynamic Time Warping (DTW). Experimental results on the UCR Archive 2018, comprising 129 time series datasets, demonstrate that employing our approach to align signals significantly enhances classification accuracy and warping average and also reduces the run time across the majority of these datasets.
- Open Access
A Low-Computational Video Synopsis Framework with a Standard Dataset
Video synopsis is an efficient method for condensing surveillance videos. This technique begins with the detection and tracking of objects, followed by the creation of object tubes. These tubes consist of sequences, each containing chronologically ordered bounding boxes of a unique object. To generate a condensed video, the first step involves rearranging the object tubes to maximize the number of non-overlapping objects in each frame. Then, these tubes are stitched to a background image extracted from the source video. The lack of a standard dataset for the video synopsis task hinders the comparison of different video synopsis models. This paper addresses this issue by introducing a standard dataset, called SynoClip, designed specifically for the video synopsis task. SynoClip includes all the necessary features needed to evaluate various models directly and effectively. Additionally, this work introduces a video synopsis model, called FGS, with low computational cost. The model includes an empty-frame object detector to identify frames empty of any objects, facilitating efficient utilization of the deep object detector. Moreover, a tube grouping algorithm is proposed to maintain relationships among tubes in the synthesized video. This is followed by a greedy tube rearrangement algorithm, which efficiently determines the start time of each tube. Finally, the proposed model is evaluated using the proposed dataset. The source code, fine-tuned object detection model, and tutorials are available at https://github.com/Ramtin-ma/VideoSynopsis-FGS.
Open AccessA Low-Computational Video Synopsis Framework with a Standard Dataset
Video synopsis is an efficient method for condensing surveillance videos. This technique begins with the detection and tracking of objects, followed by the creation of object tubes. These tubes consist of sequences, each containing chronologically ordered bounding boxes of a unique object. To generate a condensed video, the first step involves rearranging the object tubes to maximize the number of non-overlapping objects in each frame. Then, these tubes are stitched to a background image extracted from the source video. The lack of a standard dataset for the video synopsis task hinders the comparison of different video synopsis models. This paper addresses this issue by introducing a standard dataset, called SynoClip, designed specifically for the video synopsis task. SynoClip includes all the necessary features needed to evaluate various models directly and effectively. Additionally, this work introduces a video synopsis model, called FGS, with low computational cost. The model includes an empty-frame object detector to identify frames empty of any objects, facilitating efficient utilization of the deep object detector. Moreover, a tube grouping algorithm is proposed to maintain relationships among tubes in the synthesized video. This is followed by a greedy tube rearrangement algorithm, which efficiently determines the start time of each tube. Finally, the proposed model is evaluated using the proposed dataset. The source code, fine-tuned object detection model, and tutorials are available at https://github.com/Ramtin-ma/VideoSynopsis-FGS.
- Open Access
- 21 February 2024
- Published by SCI AND TECH UNIVERSAL INC in Scientia Iranica
Temporal Analysis of Functional Brain Connectivity for EEG-based Emotion Recognition
EEG signals in emotion recognition absorb special attention owing to their high temporal resolution and their information about brain activity. Different brain areas work together and the activity of brain changes over time. In this study, we investigate the emotion classification performance using functional connectivity features in different frequency bands and compare them with the classification performance using differential entropy feature, which has been previously used for this task. Moreover, we investigate the effect of different time periods on classification performance. Our results on SEED dataset show that as time goes on, emotions become more stable and the classification accuracy increases. Among different time periods, we achieve the highest classification accuracy using the time period of 140s-end. In this time period, the accuracy is improved by 4 to 6% compared to the entire signal. Pearson correlation coefficient, coherence and phase locking value features and SVM obtain the mean accuracy of about 88%. Using the proposed framework, functional connectivity features lead to better classification accuracy than DE features (with the mean accuracy of 84.89%). Finally, using the best time interval and SVM, we achieve better accuracy than using RNNs which need large amount of data and have high computational cost.
- 21 February 2024
- Published by SCI AND TECH UNIVERSAL INC in Scientia Iranica
Open AccessTemporal Analysis of Functional Brain Connectivity for EEG-based Emotion Recognition
EEG signals in emotion recognition absorb special attention owing to their high temporal resolution and their information about brain activity. Different brain areas work together and the activity of brain changes over time. In this study, we investigate the emotion classification performance using functional connectivity features in different frequency bands and compare them with the classification performance using differential entropy feature, which has been previously used for this task. Moreover, we investigate the effect of different time periods on classification performance. Our results on SEED dataset show that as time goes on, emotions become more stable and the classification accuracy increases. Among different time periods, we achieve the highest classification accuracy using the time period of 140s-end. In this time period, the accuracy is improved by 4 to 6% compared to the entire signal. Pearson correlation coefficient, coherence and phase locking value features and SVM obtain the mean accuracy of about 88%. Using the proposed framework, functional connectivity features lead to better classification accuracy than DE features (with the mean accuracy of 84.89%). Finally, using the best time interval and SVM, we achieve better accuracy than using RNNs which need large amount of data and have high computational cost.
- Open Access
- 19 October 2023
- Published by Springer Nature in Research Square
Video alignment using unsupervised learning of local and global features
In this paper, we tackle the problem of video alignment, the process of matching the frames of a pair of videos containing similar actions. The main challenge in video alignment is that accurate correspondence should be established despite the differences in the execution processes and appearances between the two videos. We introduce an unsupervised method for alignment that uses global and local features of the frames. In particular, we introduce effective features for each video frame using three machine vision tools: person detection, pose estimation, and VGG network. Then, the features are processed and combined to construct a multidimensional time series that represents the video. The resulting time series are used to align videos of the same actions using a novel version of dynamic time warping named Diagonalized Dynamic Time Warping(DDTW). The main advantage of our approach is that no training is required, which makes it applicable for any new type of action without any need to collect training samples for it. For evaluation, we considered video synchronization and phase classification tasks on the Penn action dataset \citep{penn}. Also, for an effective evaluation of the video synchronization task, we present a new metric called Enclosed Area Error(EAE). The results show that our method outperforms previous state-of-the-art methods, such as TCC \citep{TCC}, and other self-supervised and weakly supervised methods.
- 19 October 2023
- Published by Springer Nature in Research Square
Open AccessVideo alignment using unsupervised learning of local and global features
In this paper, we tackle the problem of video alignment, the process of matching the frames of a pair of videos containing similar actions. The main challenge in video alignment is that accurate correspondence should be established despite the differences in the execution processes and appearances between the two videos. We introduce an unsupervised method for alignment that uses global and local features of the frames. In particular, we introduce effective features for each video frame using three machine vision tools: person detection, pose estimation, and VGG network. Then, the features are processed and combined to construct a multidimensional time series that represents the video. The resulting time series are used to align videos of the same actions using a novel version of dynamic time warping named Diagonalized Dynamic Time Warping(DDTW). The main advantage of our approach is that no training is required, which makes it applicable for any new type of action without any need to collect training samples for it. For evaluation, we considered video synchronization and phase classification tasks on the Penn action dataset \citep{penn}. Also, for an effective evaluation of the video synchronization task, we present a new metric called Enclosed Area Error(EAE). The results show that our method outperforms previous state-of-the-art methods, such as TCC \citep{TCC}, and other self-supervised and weakly supervised methods.
- Open Access
Dynamic Changes of Brain Network during Epileptic Seizure
Epilepsy is a neurological disorder identified by sudden and recurrent seizures, which are believed to be accompanied by distinct changes in brain dynamics. Exploring the dynamic changes of brain network states during seizures can pave the way for improving the diagnosis and treatment of patients with epilepsy. In this paper, the connectivity brain network is constructed using the phase lag index (PLI) measurement within five frequency bands, and graph-theoretic techniques are employed to extract topological features from the brain network. Subsequently, an unsupervised clustering approach is used to examine the state transitions of the brain network during seizures. Our findings demonstrate that the level of brain synchrony during the seizure period is higher than the pre-seizure and post-seizure periods in the theta, alpha, and beta bands, while it decreases in the gamma bands. These changes in synchronization also lead to alterations in the topological features of functional brain networks during seizures. Additionally, our results suggest that the dynamics of the brain during seizures are more complex than the traditional three-state model (pre-seizure, seizure, and post-seizure) and the brain network state exhibits a slower rate of change during the seizure period compared to the pre-seizure and post-seizure periods.
Open AccessDynamic Changes of Brain Network during Epileptic Seizure
Epilepsy is a neurological disorder identified by sudden and recurrent seizures, which are believed to be accompanied by distinct changes in brain dynamics. Exploring the dynamic changes of brain network states during seizures can pave the way for improving the diagnosis and treatment of patients with epilepsy. In this paper, the connectivity brain network is constructed using the phase lag index (PLI) measurement within five frequency bands, and graph-theoretic techniques are employed to extract topological features from the brain network. Subsequently, an unsupervised clustering approach is used to examine the state transitions of the brain network during seizures. Our findings demonstrate that the level of brain synchrony during the seizure period is higher than the pre-seizure and post-seizure periods in the theta, alpha, and beta bands, while it decreases in the gamma bands. These changes in synchronization also lead to alterations in the topological features of functional brain networks during seizures. Additionally, our results suggest that the dynamics of the brain during seizures are more complex than the traditional three-state model (pre-seizure, seizure, and post-seizure) and the brain network state exhibits a slower rate of change during the seizure period compared to the pre-seizure and post-seizure periods.
- 8 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 23 (15) , 17751-17760
Blood Pressure Estimation Using a PPG Signal Recorded From the Fingertip of a Hand Moving in the Sagittal Plane
In this paper, we present a new photoplethysmogram (PPG)-based Blood Pressure (BP) estimation approach, in which, the PPG recording site (fingertip) is not in a standstill position, but swings up and down in the sagittal plane during the recording. The upward and downward movements of the recording site result in useful morphological changes in the PPG signal, which can be used to better estimate BP. To investigate the effect of the vertical hand movement on the shape of the PPG signal, we first devised a PPG recording hardware and used it to collect a dataset consisting of PPG signals recorded from 120 subjects. Then, we utilized machine-learning models to estimate BP using the extracted features from the recorded PPG signals. The mean absolute and the standard deviation of the error for systolic and diastolic BP estimations were 6.63±5.59 mmHg and 3.96±3.82 mmHg, respectively. The results comply with the AAMI standard for BP measurement. They also achieve Grade A for diastolic, and Grade B for systolic BP and mean BP, based on BHS standard. We conclude that the changes in PPG signal morphology provide more informative features resulting in more accurate BP estimation compared to using PPG signals recorded from a standstill finger. Moreover, the proposed BP estimation method has great potential for implementation in smartphones to monitor BP more accurately than the previous methods in which the recording site is in a standstill position.
- 8 June 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 23 (15) , 17751-17760
Blood Pressure Estimation Using a PPG Signal Recorded From the Fingertip of a Hand Moving in the Sagittal Plane
In this paper, we present a new photoplethysmogram (PPG)-based Blood Pressure (BP) estimation approach, in which, the PPG recording site (fingertip) is not in a standstill position, but swings up and down in the sagittal plane during the recording. The upward and downward movements of the recording site result in useful morphological changes in the PPG signal, which can be used to better estimate BP. To investigate the effect of the vertical hand movement on the shape of the PPG signal, we first devised a PPG recording hardware and used it to collect a dataset consisting of PPG signals recorded from 120 subjects. Then, we utilized machine-learning models to estimate BP using the extracted features from the recorded PPG signals. The mean absolute and the standard deviation of the error for systolic and diastolic BP estimations were 6.63±5.59 mmHg and 3.96±3.82 mmHg, respectively. The results comply with the AAMI standard for BP measurement. They also achieve Grade A for diastolic, and Grade B for systolic BP and mean BP, based on BHS standard. We conclude that the changes in PPG signal morphology provide more informative features resulting in more accurate BP estimation compared to using PPG signals recorded from a standstill finger. Moreover, the proposed BP estimation method has great potential for implementation in smartphones to monitor BP more accurately than the previous methods in which the recording site is in a standstill position.
- Open Access
Face: Fast, Accurate and Context-Aware Audio Annotation and Classification
This paper presents a context-aware framework for feature selection and classification procedures to realize a fast and accurate audio event annotation and classification. The context-aware design starts with exploring feature extraction techniques to find an appropriate combination to select a set resulting in remarkable classification accuracy with minimal computational effort. The exploration for feature selection also embraces an investigation of audio Tempo representation, an advantageous feature extraction method missed by previous works in the environmental audio classification research scope. The proposed annotation method considers outlier, inlier, and hard-to-predict data samples to realize context-aware Active Learning, leading to the average accuracy of 90% when only 15% of data possess initial annotation. Our proposed algorithm for sound classification obtained average prediction accuracy of 98.05% on the UrbanSound8K dataset. The notebooks containing our source codes and implementation results are available at https://github.com/gitmehrdad/FACE.
Open AccessFace: Fast, Accurate and Context-Aware Audio Annotation and Classification
This paper presents a context-aware framework for feature selection and classification procedures to realize a fast and accurate audio event annotation and classification. The context-aware design starts with exploring feature extraction techniques to find an appropriate combination to select a set resulting in remarkable classification accuracy with minimal computational effort. The exploration for feature selection also embraces an investigation of audio Tempo representation, an advantageous feature extraction method missed by previous works in the environmental audio classification research scope. The proposed annotation method considers outlier, inlier, and hard-to-predict data samples to realize context-aware Active Learning, leading to the average accuracy of 90% when only 15% of data possess initial annotation. Our proposed algorithm for sound classification obtained average prediction accuracy of 98.05% on the UrbanSound8K dataset. The notebooks containing our source codes and implementation results are available at https://github.com/gitmehrdad/FACE.