Department of Computational Systems and Applications
Head: Emmanouilidis Christos
Postgraduate researchers: Koutsiamanis Remous - Aris, Pistofidis Petros, Tasidou Aimilia
Research assistants: Papathanasiou Nikos, Pyloudis Konstantinos, Karampatzakis Dimitriοs
Research in Computational Systems and Application
In many real world problems there is a clear need for developing solutions, which are flexible, adaptive and scalable, while at the same time they are capable of exploiting recent advances in Information and Communications Technologies. Such demands can be satisfied by computational architectures and systems which utilize advanced tools and techniques, such as computational intelligence, data mining, knowledge management and in general modern ICT technologies. Similar computational systems are increasingly deployed in a wide range of contexts. In culture and education, specifically, prime targeted applications are in processing, managing and promoting cultural objects and content, digital libraries, as well as in developing flexible and adaptive learning solutions, in collaboration with the other CETI units and research labs, the neighboring Democritus University of Thrace, as well as other partners in academia and industry.
Application development is based on techniques and tools such as:
- Computational Intelligence (neural networks, fuzzy logic, evolutionary computation etc.), pattern recognition, feature selection and multi-criteria decision making.
- Digitization of objects, as well as processing and delivery of still images and video.
- Document Image Analysis
- Adaptive and personalized computing, utilizing wireless technologies.
- Wired/Wireless Multimedia Delivery
Digital Libraries / Document Image Analysis. The analysis, management and processing of digitized content, which is becoming increasingly available after large-scale digitization projects in recent years, presents considerable research challenges. There is considerable research potential in intelligent techniques and tools, which will enable to move forward from current practice in individually and manually processing such content, towards advanced environments for interactive and partially-automated processing and management of the digitized collections. Typical areas of our research focus, in collaboration with the Democritus University of Thrace, are as follows:
- Automated OCR and word spotting in document images.
- Smart page layout analysis
- Enhancement and restoration of digital document images
- Intelligent and efficient searches in digitized collections.
where current research employs intelligent techniques (neural networks, multicriteria evolutionary computing, fuzzy logic, statistical learning) depending on the specific application needs.
For further information please contact any of the team members: Christos Emmanouilidis; Costas Mpatsalas, Sotiris Delis.
Personalized and context-aware computing. The increasing spread of mobile and embedded computing devices, coupled with the deeper penetration of broadband and wireless networking requires focused research efforts in developing applications and solutions which can be adapted to serve diverse device and user needs. The emergence of personalized and contextualized services and applications, which can be deployed in heterogeneous environments and can be adjusted to specific user needs, is part of the new breed of innovative computing solutions, which are expected to achieve deep penetration in various application sectors.
Typical examples are multi-platform (desktop, PDA, laptop, tablet, touch-screen) deployable applications, which can offer ubiquitous services to users (anyone, anytime, anywhere), while they are capable of adapting the delivered content based on location, user or context awareness. For example the system under development in the Virtual Guide project is suitable for deployment primarily in museums and exhibition areas and will facilitate visitor navigation and enhance its experience via the use of wireless handheld devices, which exploit location awareness features in order to deliver semantically relevant content targeted to specific user profiles. The system is capable of handling and delivering multimedia-rich information (textual, audio, images, video), taking into account the visitors location and their specific profiles. The CETI Cultural Heritage Unit is also heavily involved in this project, which, is a collaboration between CETI, PRISMA Electronics SA, the South-eastern Europe Telecommunications and Informatics Research Institute (INA), the Alexandroupolis Ecclesiastical Museum, while there is also support from the Democritus University of Thrace.
For further information please contact any of the team members: Christos Emmanouilidis; Stelios Demitriou, Eleni Kamateri, Alexandros Michalakidis and Panagiotis Papademetriou.