A5047979812- Advanced Image and Video Retrieval Techniques
- EEG and Brain-Computer Interfaces
- ECG Monitoring and Analysis
- Image Retrieval and Classification Techniques
- Neural Networks and Applications
- COVID-19 diagnosis using AI
- Non-Invasive Vital Sign Monitoring
- Music and Audio Processing
- Remote-Sensing Image Classification
- Image and Signal Denoising Methods
- Video Analysis and Summarization
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Machine Learning and ELM
- Speech and Audio Processing
- Blind Source Separation Techniques
- Machine Fault Diagnosis Techniques
- Metaheuristic Optimization Algorithms Research
- Structural Health Monitoring Techniques
- Radiomics and Machine Learning in Medical Imaging
- Anomaly Detection Techniques and Applications
- AI in cancer detection
- Infrastructure Maintenance and Monitoring
- Face and Expression Recognition
- Soil Moisture and Remote Sensing
- Advanced SAR Imaging Techniques
Qatar University
2015-2024
IRD Fuel Cells (Denmark)
2024
Tampere University
2009-2021
İzmir University of Economics
2021
Signal Processing (United States)
2003-2017
Tampere University of Applied Sciences
2004-2015
Middle East Technical University
2013
Research Council of Finland
2009
Nokia (Finland)
2000-2002
During the last decade, Convolutional Neural Networks (CNNs) have become de facto standard for various Computer Vision and Machine Learning operations. CNNs are feed-forward Artificial (ANNs) with alternating convolutional subsampling layers. Deep 2D many hidden layers millions of parameters ability to learn complex objects patterns providing that they can be trained on a massive size visual database ground-truth labels. With proper training, this unique makes them primary tool engineering...
This paper presents a fast and accurate patient-specific electrocardiogram (ECG) classification monitoring system.An adaptive implementation of 1-D convolutional neural networks (CNNs) is inherently used to fuse the two major blocks ECG into single learning body: feature extraction classification. Therefore, for each patient, an individual simple CNN will be trained by using relatively small common training data, thus, such ability can further improve performance. Since this also negates...
Early detection of the motor faults is essential and artificial neural networks are widely used for this purpose. The typical systems usually encapsulate two distinct blocks: feature extraction classification. Such fixed hand-crafted features may be a suboptimal choice require significant computational cost that will prevent their usage real-time applications. In paper, we propose fast accurate condition monitoring early fault-detection system using 1-D convolutional has an inherent adaptive...
This paper presents a generic and patient-specific classification system designed for robust accurate detection of ECG heartbeat patterns. The proposed feature extraction process utilizes morphological wavelet transform features, which are projected onto lower dimensional space using principal component analysis, temporal features from the data. For pattern recognition unit, feedforward fully connected artificial neural networks, optimally each patient by multidimensional particle swarm...
1D Convolutional Neural Networks (CNNs) have recently become the state-of-the-art technique for crucial signal processing applications such as patient-specific ECG classification, structural health monitoring, anomaly detection in power electronics circuitry and motor-fault detection. This is an expected outcome there are numerous advantages of using adaptive compact CNN instead a conventional (2D) deep counterparts. First all, CNNs can be efficiently trained with limited dataset signals...
Automated early detection and identification of switch faults are essential in high-voltage applications. Modular multilevel converter (MMC) is a new promising topology for such MMC composed many identical controlled voltage sources called modules or cells. Each cell may have one more switches failure occur anyone these The steady-state normal fault behavior will also significantly vary according to the changes load current timing. This makes it challenging problem detect identify as soon...
We propose a fast and accurate patient-specific electrocardiogram (ECG) classification monitoring system using an adaptive implementation of 1D Convolutional Neural Networks (CNNs) that can fuse feature extraction into unified learner. In this way, dedicated CNN will be trained for each patient by relatively small common training data thus it also used to classify long ECG records such as Holter registers in manner. Alternatively, solution conveniently real-time early alert on light-weight...
The immense spread of coronavirus disease 2019 (COVID-19) has left healthcare systems incapable to diagnose and test patients at the required rate. Given effects COVID-19 on pulmonary tissues, chest radiographic imaging become a necessity for screening monitoring disease. Numerous studies have proposed Deep Learning approaches automatic diagnosis COVID-19. Although these methods achieved outstanding performance in detection, they used limited X-ray (CXR) repositories evaluation, usually with...
Detecting COVID-19 at an early stage is essential to reduce the mortality risk of patients. In this study, a cascaded system proposed segment lung, detect, localize, and quantify infections from computed tomography images. An extensive set experiments were performed using Encoder–Decoder Convolutional Neural Networks (ED-CNNs), UNet, Feature Pyramid Network (FPN), with different backbone (encoder) structures variants DenseNet ResNet. The conducted for lung region segmentation showed Dice...
Coronavirus disease (COVID-19) has been the main agenda of whole world ever since it came into sight. X-ray imaging is a common and easily accessible tool that great potential for COVID-19 diagnosis prognosis. Deep learning techniques can generally provide state-of-the-art performance in many classification tasks when trained properly over large data sets. However, scarcity be crucial obstacle using them detection. Alternative approaches such as representation-based [collaborative or sparse...
Diabetes foot ulceration (DFU) and amputation are a cause of significant morbidity. The prevention DFU may be achieved by the identification patients at risk institution preventative measures through education offloading. Several studies have reported that thermogram images help to detect an increase in plantar temperature prior DFU. However, distribution heterogeneous, making it difficult quantify utilize predict outcomes. We compared machine learning-based scoring technique with feature...
Polarimetric synthetic aperture radar (PolSAR) data are used extensively for terrain classification applying SAR features from various target decompositions and certain textural features. However, one source of information has so far been neglected PolSAR classification: Color. It is a common practice to visualize by color coding methods thus, it possible extract powerful such pseudocolor images as provide additional superior classification. In this paper, we first review previous attempts...
Phonocardiogram * (PCG) signal is used as a diagnostic test in ambulatory monitoring order to evaluate the heart hemodynamic status and detect cardiovascular disease.The objective of this study develop an automatic classification method for anomaly (normal vs. abnormal) quality (good bad) detection PCG recordings without segmentation.For purpose, subset 18 features selected among 40 based on wrapper feature selection scheme.These are extracted from time, frequency, time-frequency domains any...
1 Atrial Fibrillation (AF) is characterized by chaotic electrical impulses in the atria, which leads to irregular heartbeats and can develop blood clots stroke.Therefore, early detection of AF crucial for increasing success rate treatment.This study focused on rhythm using hand-held ECG monitoring devices, addition three other classes: normal or sinus rhythm, rhythms, too noisy analyze.The pipeline proposed method consists major components: preprocessing feature extraction, selection,...
In this paper, the performance of phase space representation in interpreting underlying dynamics epileptic seizures is investigated and a novel patient-specific seizure detection approach proposed based on EEG signals. To accomplish this, trajectories nonseizure segments are reconstructed high dimensional using time-delay embedding method. Afterwards, Principal Component Analysis (PCA) was used order to reduce dimension spaces. The geometry lower dimensions then characterized Poincaré...