A5036751601- Hand Gesture Recognition Systems
- Human Pose and Action Recognition
- Gait Recognition and Analysis
- Hearing Impairment and Communication
- Blood Pressure and Hypertension Studies
- Automated Road and Building Extraction
- Remote Sensing and Land Use
- Nutritional Studies and Diet
- Remote-Sensing Image Classification
- Cardiovascular Disease and Adiposity
- Horticultural and Viticultural Research
- Advanced Vision and Imaging
- Anomaly Detection Techniques and Applications
- Advanced Neural Network Applications
- Remote Sensing and LiDAR Applications
- Heart Rate Variability and Autonomic Control
- Healthcare Systems and Public Health
- Occupational Health and Safety Research
- Industrial Vision Systems and Defect Detection
- Diet and metabolism studies
- Atrial Fibrillation Management and Outcomes
- linguistics and terminology studies
- Video Surveillance and Tracking Methods
- Robot Manipulation and Learning
- Musculoskeletal pain and rehabilitation
National and Kapodistrian University of Athens
2022-2025
Hippocration General Hospital
2023-2025
Information Technologies Institute
2018-2024
Centre for Research and Technology Hellas
2018-2024
Ippokrateio General Hospital of Thessaloniki
2024
Weatherford College
2022
AHEPA University Hospital
2022
Aristotle University of Thessaloniki
2022
Kingston University
2019
China Philanthropy Research Institute
2018
Sign language recognition (SLR) is a challenging, but highly important research field for several computer vision systems that attempt to facilitate the communication among deaf and hearing impaired people. In this work, we propose an accurate robust deep learning-based methodology sign from video sequences. Our novel method relies on hand body skeletal features extracted RGB videos and, therefore, it acquires discriminative gesture data without need any additional equipment, such as gloves,...
In recent years, major advances in artificial intelligence (AI) have led to the development of powerful AI systems for use field nutrition order enhance personalized dietary recommendations and improve overall health well-being. However, lack guidelines from nutritional experts has raised questions on accuracy trustworthiness advice provided by such systems. This paper aims address this issue introducing a novel AI-based recommendation method that leverages speed explainability deep...
Continuous Sign Language Recognition (CSLR) refers to the challenging problem of recognizing sign language glosses and their temporal boundaries from weakly annotated video sequences. Previous methods focus mostly on visual feature extraction neglecting text information failing effectively model intra-gloss dependencies. In this work, a cross-modal learning approach that leverages improve vision-based CSLR is proposed. To end, two powerful encoding networks are initially used produce...
Building detection from two-dimensional high-resolution satellite images is a computer vision, photogrammetry, and remote sensing task that has arisen in the last decades with advances sensors technology can be utilized several applications require creation of urban maps or study changes. However, variety irrelevant objects appear an environment resemble buildings, significant variations shape generally appearance buildings render building quite demanding task. As result, automated methods...
Sign language recognition (SLR) refers to the classification of signs with a specific meaning performed by deaf and/or hearing-impaired people in their everyday communication. In this work, we propose deep learning based framework, which examine and analyze contribution video (image optical flow) skeletal (body, hand face) features challenging task isolated SLR, each signed corresponds single word. Moreover, employ various fusion schemes order identify optimal way combine information...
The field of 3D hand pose estimation has been gaining a lot attention recently, due to its significance in several applications that require human-computer interaction (HCI). utilization technological advances, such as cost-efficient depth cameras coupled with the explosive progress Deep Neural Networks (DNNs), led significant boost development robust markerless methods. Nonetheless, finger occlusions and rapid motions still challenges accuracy In this survey, we provide comprehensive study...
AI-based software applications for personalized nutrition have recently gained increasing attention to help users follow a healthy lifestyle. In this paper, we present knowledge-based recommendation framework that exploits an explicit dataset of expert-validated meals offer highly accurate diet plans spanning across ten user groups both subjects and participants with health conditions. The proposed advisor is built on novel architecture includes (a) qualitative layer verifying ingredient...
Abstract Hypertension remains a major public health challenge with inadequate control globally. The May Measurement Month (MMM) global survey initiated by the International Society of (ISH) was implemented in Greece 2022 aiming to raise hypertension awareness and control. Adult volunteers were recruited through opportunistic screening 11 urban areas. Information on medical history three sitting blood pressure (BP) measurements obtained using validated automated upper-arm devices data...
A significant proportion of hypertensive patients are both smokers and obese. Several pathophysiological mechanisms involved in the combined effect smoking obesity on hypertension onset maintenance. These include increased sympathetic nervous system activity, endothelial dysfunction, inflammation, oxidative stress, insulin resistance. The presence these major cardiovascular risk factors may lead to difficult-to-control as well substantially increase for an adverse outcome. It is, therefore,...
Vision Transformers are very popular nowadays due to their state-of-the-art performance in several computer vision tasks, such as image classification and action recognition. Although has been greatly enhanced through highly descriptive patch embeddings hierarchical structures, there is still limited research on utilizing additional data representations so refine the self-attention map of a Transformer. To address this problem, novel attention mechanism, called multi-manifold multi-head...
Grape maturity estimation is vital in precise agriculture as it enables informed decision making for disease control, harvest timing, grape quality, and quantity assurance. Despite its importance, there are few large publicly available datasets that can be used to train accurate robust segmentation algorithms. To this end, work proposes the CERTH dataset, a new sizeable dataset designed explicitly evaluating deep learning algorithms estimation. The proposed one of largest currently...
Eating behavior can have an important effect on, and be correlated with, obesity eating disorders. is usually estimated through self-reporting measures, despite their limitations in reliability, based on ease of collection analysis. A better widely used alternative the objective analysis during meals human annotations in-meal behavioral events (e.g., bites). However, this methodology time-consuming often affected by error, limiting its scalability cost-effectiveness for large-scale research....
Ergonomic risk assessment is vital for identifying work-related human postures that can be detrimental to the health of a worker. Traditionally, ergonomic risks are reported by experts through time-consuming and error-prone procedures; however, automatic algorithmic methods have recently started emerge. To further facilitate assessment, this paper proposes novel variational deep learning architecture estimate any task utilizing Rapid Entire Body Assessment (REBA) framework. The proposed...
Phase correlation (PC) is widely employed by several sub-pixel motion estimation techniques in an attempt to accurately and robustly detect the displacement between two images. To achieve accuracy, these employ interpolation methods function-fitting approaches on cross-correlation function derived from PC core. However, such still present a lower bound of accuracy that cannot be overcome. allow room for further improvements, we propose this paper enhancement employing completely different...
Ahstract- The accuracy of modern depth sensors, the robustness skeletal data to illumination variations and superb performance deep learning techniques on several classification tasks have sparkled a renewed intereste towards skeleton-based action recognition. In this paper, we propose four-stream neural network based two types spatial features their corresponding temporal representations extracted by novel Grassmannian Pyramid Descriptor (GPD). proposed recognition methodology is further...