CERTH
Description
The Centre for Research and Technology-Hellas (CERTH), founded in 2000, is the only research centre in Northern Greece and one of the largest in the country, and consists of the five institutes: (a) Chemical Process & Energy Resources Institute (CPERI), (b) Information Technologies Institute (ITI), (c) Hellenic Institute of Transport (HIT), (d) Institute of Applied Biosciences (INAB), (e) Institute for Research and Technology Thessaly (IRETETH). In 7SHIELD, CERTH participates through ITI.
The Information Technologies Institute (ITI) of CERTH has been a founding member of CERTH since 2000.
The participating team, namely the Multimodal Data Fusion and Analytics Group (M4D) of the Multimedia Knowledge and Social Data Analytics laboratory (MKLab), has significant security research experience and scientific expertise on artificial intelligence and more specifically in the fields of machine learning, deep learning, multimodal retrieval, multimedia analysis, visual analytics and decision support systems, computer vision, knowledge engineering, semantic integration of heterogeneous resources, semantic reasoning, discovery and mining of heterogeneous multilingual and multimedia Web resources, social media monitoring, as well as on the processing, analysis of the multimodal data extracted from them. In recent years, the team has coordinated and participated in more than 150 European and National research projects in the areas of multimedia processing, information extraction, and social media monitoring and analysis.
Role in the project
CERTH is responsible for the management of scientific and technical aspects of the 7SHIELD project and has a critical role in the project by leading a WP for the detection cyber-physical attacks and tasks including those to respond in emergency situations. The developed solutions cover computer vision analysis, ontological modeling and crisis classification for risk assessment in ground segments of space systems.
Face detection and face recognition module (KR06)
State-of-the-art methodologies for optical video surveillance to recognize human malicious activities, detect and identify faces from various surveillance cameras will be incorporated into 7SHIELD framework aiming to strengthen the 7SHIELD’s physical and cyber detection toolset.
Video-based object and activity recognition module (KR07)
Innovative methodologies for the object detection and the activity recognition services via processing video streams from the surveillance area will be adopted in the 7SHIELD P/C detection toolset. Specifically, the Video-based object and activity recognition module will process video streams or still images in order to locate and recognize objects of interest in the provided sources. Additionally, after detecting any human presence in the scene the corresponding results of object detection will be propagated to the activity recognition sub-module to identify suspicious and harmful activities.
7SHIELD Knowledge Base (KR12)
State-of-the-art methodologies for semantic representation and linking for reasoning so as to represent data in semantic format, save semantic models, enrich using reasoning mechanisms and retrieve using semantic queries will be encompassed into 7SHIELD framework. The aim is to build an ontological framework which will be capable of representing each kind of information coming from the outputs of the 7SHIELD tools.
Crisis classification module (KR13)
The accurate and timely estimation of the severity of the crisis is an ultimate goal for authorities to effectively respond and handle the ongoing crisis. 7SHIELD aims to encompass methodologies for multi-level real-time P/C crisis assessments that rely on multimodal information and data fusion. Specifically, the Crisis Classification module will analyse multiple types of data, which are generated as outcomes of the detection tools for P/C threats, classify crisis events by utilising machine learning techniques, supporting the decision-making processes.