The contemporary world is becoming increasingly dependent on automated computer systems, utilized in various sectors. Reliable Cybersecurity solutions have become essential towards protecting these systems, which are susceptible to cyber attacks aiming at the disclosure or modification of information, and in some cases at the complete disruption of their operation. Deep learning architectures offer novel solutions in this regard, providing methods for detecting and possibly mitigating such threats.

​ The aim of this competition is to challenge participants to develop robust models capable of performing anomaly detection, classification, and attack recognition in two Cybersecurity applications. In the first case, which targets large power networks, participants will have to develop a reliable detector, able to identify with high precision anomalies in the provided network traffic dataset.

​The second case involves environmental attacks on autonomous vehicles, based on modifications in traffic signs that aim to confuse the vehicle, or guide it to a different direction. The developed model in this case should be able to recognize these attacks on the given dataset of traffic sign images and classify them to the correct category. 

For more information: https://competition.ioti4-2021.web.uowm.gr/