- Modular Robots and Swarm Intelligence
- Human Pose and Action Recognition
- Smart Agriculture and AI
- Video Surveillance and Tracking Methods
- Cellular Automata and Applications
- Anomaly Detection Techniques and Applications
- Robotic Path Planning Algorithms
- Remote Sensing and LiDAR Applications
- Advanced Neural Network Applications
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Advanced Vision and Imaging
- Robotics and Sensor-Based Localization
- Image and Signal Denoising Methods
- Infrastructure Maintenance and Monitoring
- Advanced Image Processing Techniques
- Maritime Navigation and Safety
- Marine and coastal ecosystems
- Oil Spill Detection and Mitigation
- Generative Adversarial Networks and Image Synthesis
- Context-Aware Activity Recognition Systems
- Distributed Control Multi-Agent Systems
- Image Processing Techniques and Applications
- Advanced Image Fusion Techniques
- Vehicle License Plate Recognition
Information Technologies Institute
2018-2025
Centre for Research and Technology Hellas
2018-2025
Southern Health NHS Foundation Trust
2024
University of Southampton
2024
China Philanthropy Research Institute
2019-2022
UPMC Center for High Value Health Care
2022
Queen Mary University of London
2021
Democritus University of Thrace
2010-2019
Cambridgeshire and Peterborough NHS Foundation Trust
2016
Addenbrooke's Hospital
2016
Oil spill is considered one of the main threats to marine and coastal environments. Efficient monitoring early identification oil slicks are vital for corresponding authorities react expediently, confine environmental pollution avoid further damage. Synthetic aperture radar (SAR) sensors commonly used this objective due their capability operating efficiently regardless weather illumination conditions. Black spots probably related spills can be clearly captured by SAR sensors, yet...
In this article, the problem of real-time robot exploration and map building (active SLAM) is considered. A single stereo vision camera exploited by a fully autonomous to navigate, localize itself, define its surroundings, avoid any possible obstacle in aim maximizing mapped region following optimal route. modified version so-called cognitive-based adaptive optimization algorithm introduced for successfully complete tasks real time local minima entrapment. The method's effectiveness...
The past few years have seen an accelerating integration of deep learning (DL) techniques into various remote sensing (RS) applications, highlighting their power to adapt and achieving unprecedented advancements. In the present review, we provide exhaustive exploration DL approaches proposed specifically for spatial downscaling RS imagery. A key contribution our work is presentation major architectural components models, metrics, data sets available this task as well construction a compact...
The CoFly-WeedDB contains 201 RGB images (∼436 MB) from the attached camera of DJI Phantom Pro 4 a cotton field in Larissa, Greece during first stages plant growth. 1280 × 720 were collected while Unmanned Aerial Vehicle (UAV) was performing coverage mission over field's area. During designed mission, angle adjusted to -87°, vertically with field. flight altitude and speed UAV equal 5 m 3 m/s, respectively, aiming provide close clear view weed instances. All have been annotated by expert...
Abstract This paper presents a novel, low-cost, user-friendly Precision Agriculture platform that attempts to alleviate the drawbacks of limited battery life by carefully designing missions tailored each field’s specific, time-changing characteristics. The proposed system is capable coverage for any type UAV, integrating field characteristics into resulting trajectory, such as irregular shape and obstacles. collected images are automatically processed create detailed orthomosaics extract...
Gas cylinder detection and the identification of their characteristics hold considerable potential for enhancing safety operational efficiency in several applications, including industrial warehouse operations. These tasks gain significance with growth online trade, emerging as critical instruments to combat environmental crimes associated hazardous substances' illegal commerce. However, lack relevant datasets hinders effective utilization deep learning techniques within this domain. In...
Oil spills pose a major threat of the oceanic and coastal environments, hence, an automatic detection continuous monitoring system comprises appealing option for minimizing response time relevant operations. Numerous efforts have been conducted towards such solutions by exploiting variety sensing systems as satellite Synthetic Aperture Radar (SAR) which can identify oil over sea surfaces in any environmental conditions operational time. Such approaches include use artificial neural networks...
Social media play an important role in the daily life of people around globe and users have emerged as active part news distribution well production. The threatening pandemic COVID-19 has been lead subject online discussions posts, resulting to large amounts related social data, which can be utilised reinforce crisis management several ways. Towards this direction, we propose a novel framework collect, analyse, visualise Twitter tailored specifically monitor virus spread severely affected...
This paper presents a modular and holistic Precision Agriculture platform, named CoFly, incorporating custom-developed AI ICT technologies with pioneering functionalities in UAV-agnostic system. Cognitional operations of micro Flying vehicles are utilized for data acquisition advanced coverage path planning obstacle avoidance functionalities. Photogrammetric outcomes extracted by processing UAV into 2D fields crop health maps, enabling the extraction high-level semantic information about...
A Cellular Automaton-based technique suitable for solving the path planning problem in a distributed robot team is outlined. Real-time challenging task that has many applications fields of artificial intelligence, moving robots, virtual reality, and agent behavior simulation. The refers to finding collision-free autonomous robots between two specified positions configuration area. complexity increases systems multiple robots. More specifically, some distance should be covered by each an...
This work examines violence detection in video scenes of crowds and proposes a crowd framework based on 3D convolutional deep learning architecture, the 3D-ResNet model with 50 layers. The proposed is evaluated Violent Flows dataset against several state-of-the-art approaches achieves higher accuracy values almost all cases, while also performing activities (near) real-time.
In this paper, an innovative technique for digital image stabilization (DIS) based on the Hilbert-Huang transform (HHT) is proposed. It exploits basic features of HHT in order to separate local motion signal obtained from sequence into two different vectors. A variety embedded systems equipped with a sensor, such as handheld cameras, mobile phones, and robots, can produce sequences observed caused by types movements: smooth camera (intentional) unwanted shaking (jitter). The has been...
In the construction domain, Digital twins are mostly used for facilities management of buildings, but their applications still very limited. The virtualization buildings and bridges in last 15 years form Building or Bridge Information Models is clearly identified as starting point DTs. industry has erected a frame with semantically rich 3D reference models that now heavily enriched visual sensor data captured on sites. This article provides an overview research current practices computer...
This article presents a case study with various developments of digital twinning sample load tests performed on several railways bridges. The is located in Extremadura, South Western Spain and its aim the generation validated, multi-layered information construct form twin as result test. conceived, not only to verify assumptions design bridge but also, optimize future maintenance plans network. particular framed within vaster European effort digitization construction sector. Research...
The timely and efficient cooperation across sectors borders during maritime crises is paramount for the safety of human lives. Maritime monitoring authorities are now realizing grave importance cross-sector cross-border information sharing. However, this compromised by diversity existing systems vast volumes heterogeneous data generated exchanged operations. In order to address these challenges, EU has been driving several initiatives, including EU-funded projects, facilitating exchange...
This work presents a spatio-temporal activity detection and recognition framework for untrimmed surveillance videos consisting of three-step pipeline: object detection, tracking, recognition. The relies on the YOLO v4 architecture Euclidean distance while recognizer uses 3D Convolutional Deep learning employing boundaries addressing it as multi-label classification. evaluation experiments VIRAT dataset achieve accurate detections temporal recognitions activities in videos, with better...
Surveillance systems currently deploy a variety of devices that can capture visual content (such as CCTV, body-worn cameras, and smartphone cameras), thus rendering the monitoring video footage obtained from multiple such complex task. This becomes especially challenging when social events involve large crowds, particularly there is risk crowd violence. paper presents demonstrates violence detection system process, analyze, alert potential stakeholders, violence-related identified in...