A5044320536- Network Security and Intrusion Detection
- Smart Grid Security and Resilience
- Anomaly Detection Techniques and Applications
- Industrial Vision Systems and Defect Detection
- Fault Detection and Control Systems
- Software-Defined Networks and 5G
- Mineral Processing and Grinding
- Music and Audio Processing
- Adversarial Robustness in Machine Learning
- Music Technology and Sound Studies
- Advanced Data and IoT Technologies
- Artificial Intelligence in Healthcare and Education
- Advanced X-ray and CT Imaging
- Imbalanced Data Classification Techniques
- Advanced machining processes and optimization
- Software Engineering Research
- Electricity Theft Detection Techniques
- Explainable Artificial Intelligence (XAI)
- Digital and Cyber Forensics
- Advanced Malware Detection Techniques
- Welding Techniques and Residual Stresses
- Machine Fault Diagnosis Techniques
University of Western Macedonia
2021-2024
Information Technologies Institute
2020-2024
Centre for Research and Technology Hellas
2021-2024
The technological leap of smart technologies and the Internet Things has advanced conventional model electrical power energy systems into a new digital era, widely known as Smart Grid. advent Grids provides multiple benefits, such self-monitoring, self-healing pervasive control. However, it also raises crucial cybersecurity privacy concerns that can lead to devastating consequences, including cascading effects with other critical infrastructures or even fatal accidents. This paper introduces...
<ns4:p>Background Anomaly detection is vital in industrial settings for identifying abnormal behaviors that suggest faults or malfunctions. Artificial intelligence (AI) offers significant potential to assist humans addressing these challenges. Methods This study compares the performance of supervised and unsupervised machine learning (ML) techniques anomaly detection. Additionally, model-specific explainability methods were employed interpret outputs. A novel approach, MLW-XAttentIon, based...
Anomaly detection is a crucial task in the industrial field, by identifying abnormal behaviors that could suggest faults or malfunctions on production line. The emergence and rapid evolution of artificial intelligence can be utilized to assist humans when dealing with such tasks. In this paper, comparative study presented regarding performance aptitude both supervised unsupervised machine learning techniques identify anomalous behaviors. Moreover, model-specific explainability methods were...
Software Defined Networking (SDN) is an innovative technology, which can be applied in a plethora of applications and areas. Recently, SDN has been identified as one the most promising solutions for industrial well. The key features include decoupling control plane from data programmability network through application development. Researchers are looking at these order to enhance Quality Service (QoS) provisioning modern applications. To this end, following work presents development...
State-of-the art Text-To-Music (TTM) generative AI models are large and require desktop or server class compute, making them infeasible for deployment on mobile phones. This paper presents an analysis of trade-offs between model compression generation performance TTM models. We study through knowledge distillation specific modifications that enable applicability over the various components (encoder, decoder). Leveraging these methods we create TinyTTM (89.2M params) achieves a FAD 3.66 KL...
State-of-the art Text-To-Music (TTM) generative AI models are large and require desktop or server class compute, making them infeasible for deployment on mobile phones. This paper presents an analysis of trade-offs between model compression generation performance TTM models. We study through knowledge distillation specific modifications that enable applicability over the various components (encoder, decoder). Leveraging these methods we create TinyTTM (89.2M params) achieves a FAD 3.66 KL...
In this work a comparative study among the known fault detection techniques Local Outlier Factor and Isolation Forest as well proposed methodology called Standardised Mahalanobis Distance is presented. The focusing on challenging problem of bearings rotating machines using vibration sensors' data. During first phase experiments, all models are applied evaluated cross-validation dataset created in lab by obtaining signals machine. second phase, outlier including one, popular, public dataset....
Defect detection is one of the core problems’ categories that Industry 4.0 concepts like automation, IoT, digitization and AI aimed to solve. In this work, a platform extends aforementioned with ones coming from 5.0 reconciliation collaboration between humans machines introduced. The proposed provides defect localization services for hard metal industry by extending current solutions exponentially growing technologies such as interpretable explainable (XAI) services, human-in-the-loop...