A5022692056- Complexity and Algorithms in Graphs
- Advanced Graph Theory Research
- Game Theory and Voting Systems
- Geographic Information Systems Studies
- Data Management and Algorithms
- Fire effects on ecosystems
- Remote Sensing in Agriculture
- VLSI and FPGA Design Techniques
- Scientific Computing and Data Management
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Computational Geometry and Mesh Generation
- Remote-Sensing Image Classification
- Graph Theory and Algorithms
- Fire Detection and Safety Systems
- Semantic Web and Ontologies
- Anomaly Detection Techniques and Applications
- Complex Network Analysis Techniques
- Seismology and Earthquake Studies
- Species Distribution and Climate Change
- Advanced Image and Video Retrieval Techniques
- IoT and Edge/Fog Computing
- Advanced Graph Neural Networks
- Data Stream Mining Techniques
- Machine Learning and Algorithms
- Advanced Database Systems and Queries
Harokopio University of Athens
2016-2025
National and Kapodistrian University of Athens
2011-2024
Max Planck Institute for Informatics
2004-2009
Institut national de recherche en informatique et en automatique
2008
Max Planck Society
2004-2007
Mathematical software and graph-theoretical algorithmic packages to efficiently model, analyze, query graphs are crucial in an era where large-scale spatial, societal, economic network data abundantly available. One such package is JGraphT, a programming library that contains very efficient generic graph structures along with large collection of state-of-the-art algorithms. The written Java stability, interoperability, performance mind. A distinctive feature this its ability model vertices...
The availability of the sheer volume Copernicus Sentinel-2 imagery has created new opportunities for exploiting deep learning methods land use cover (LULC) image classification at large scales. However, an extensive set benchmark experiments is currently lacking, i.e. models tested on same dataset, with a common and consistent metrics, in hardware. In this work, we BigEarthNet multispectral dataset to first time different state-of-the-art multi-label, multi-class LULC problem, contributing...
Suppose that each member of a set A applicants ranks subset P posts in an order preference, possibly involving ties. matching is (applicant, post) pairs such applicant and post appears at most one pair. rank-maximal which the maximum possible number are matched to their first choice post, subject condition, second so on. This relevant concept any practical situation it was studied by Irving [2003].We give algorithm compute with running time O (min( n + C , √ ) m ), where maximal rank edge...
The combat against fake news and disinformation is an ongoing, multi-faceted task for researchers in social media networks domains, which comprises not only the detection of false facts published content but also accountability mechanisms that keep a record trustfulness sources generate and, lately, deliberately distribute information. In direction detecting handling organized networks, major networking sites are currently developing strategies to block such attempts. role machine learning...
With climate change expected to exacerbate fire weather conditions, the accurate and timely anticipation of wildfires becomes increasingly crucial for disaster mitigation. In this study, we utilize SeasFire, a comprehensive global wildfire dataset with climate, vegetation, oceanic indices, human-related variables, enable seasonal forecasting machine learning. For predictive analysis, present FireCastNet, novel architecture which combines 3D convolutional encoder GraphCast, originally...
The continuous operation of Earth-orbiting satellites generates vast and ever-growing archives Remote Sensing (RS) images. Natural language presents an intuitive interface for accessing, querying, interpreting the data from such archives. However, existing Vision-Language Models (VLMs) are predominantly trained on web-scraped, noisy image-text data, exhibiting limited exposure to specialized domain RS. This deficiency results in poor performance RS-specific tasks, as commonly used datasets...
Big Earth-observation (EO) data that are made freely available by space agencies come from various archives. Therefore, users trying to develop an application need search within these archives, discover the needed data, and integrate them into their application. In this article, we argue if EO published using linked paradigm, then discovery, integration, development of applications becomes easier. We present life cycle big, linked, open show how support stages software stack developed...
This paper discusses the perspective of H2020 TEACHING project on next generation autonomous applications running in a distributed and highly heterogeneous environment comprising both virtual physical resources spanning edge-cloud continuum. puts forward human-centred vision leveraging physiological, emotional, cognitive state users as driver for adaptation optimization applications. It does so by building distributed, embedded federated learning system complemented methods tools to enforce...
The detection of early signs volcanic unrest preceding an eruption, in the form ground deformation Interferometric Synthetic Aperture Radar (InSAR) data is critical for assessing hazard. In this work we treat as a binary classification problem InSAR images, and propose novel deep learning methodology that exploits rich source synthetically generated interferograms to train quality classifiers perform equally well real interferograms. imbalanced nature problem, with orders magnitude fewer...
The detection of organised disinformation campaigns that spread fake news, by first camouflaging them as real ones is crucial in the battle against misinformation and social media. This article presents a method for classifying diffusion graphs news formed media, taking into account profiles users participate graph, their relations way spread, ignoring actual text content or messages it. increases robustness widens its applicability different contexts. results this study show proposed...
Wildfires are increasingly exacerbated as a result of climate change, necessitating advanced proactive measures for effective mitigation. It is important to forecast wildfires weeks and months in advance plan forest fuel management, resource procurement allocation. To achieve such accurate long-term forecasts at global scale, it crucial employ models that account the Earth system's inherent spatio-temporal interactions, memory effects teleconnections. We propose teleconnection-driven vision...
We consider the problem of computing exact or approximate minimum cycle bases an undirected (or directed) graph G with m edges, n vertices and nonnegative edge weights. In this problem, a {0, 1} (−1,0,1}) incidence vector is associated each space over F 2 (Q) generated by these vectors . A set cycles called basis if it forms for its space. where sum weights Cycle low weight are useful in number contexts, example, analysis electrical networks, structural engineering, chemistry, surface...
In this paper, we consider the problem of computing a minimum cycle basis an undirected graph G = ( V , E ) with n vertices and m edges. We describe efficient implementation O 3 + mn 2 log algorithm. For sparse graphs, is currently best-known This algorithm's running time can be partitioned into two parts ), respectively. Our experimental findings imply that for random graphs true bottleneck sophisticated part. A straightforward would require Ω( nm shortest-path computations. Thus, develop...
We consider the problem of computing a minimum cycle basis an undirected non-negative edge-weighted graph G with m edges and n vertices. In this problem, {0,1} incidence vector is associated each space over $\mathbb{F}_{2}$ generated by these vectors G. A set cycles called if it forms for its space. where sum weights Minimum are useful in number contexts, e.g. analysis electrical networks structural engineering. The previous best algorithm has running time O(m ω n), exponent matrix...
Ground deformation measured from interferometric synthetic aperture radar (InSAR) data is considered a sign of volcanic unrest, statistically linked to eruption. Recent studies have shown the potential using Sentinel-1 InSAR and supervised deep learning (DL) methods for detection signals, toward global hazard mitigation. However, accuracy compromised lack labeled class imbalance. To overcome this, are typically used fine-tuning DL models pretrained on ImageNet dataset. This approach suffers...
An instance of the stable marriage problem is an undirected bipartite graph G = ( X ∪ W , E ) with linearly ordered adjacency lists ties allowed in ordering. A matching M a set edges, no two which share endpoint. edge e b ∈ ∖ blocking for if either unmatched or strictly prefers to its partner and unmatched, indifferent between them. strongly there respect it. We give O nm algorithm computing matchings, where n number vertices m edges. The previous best had running time 2 ). also study this...