Samuele Tomassacci

Information

Bio

I am Samuele Tomassacci, a student with a Bachelor's degree in Computer Engineering from the University of Rome Tor Vergata. I later specialized in Artificial Intelligence and Robotics during my Master's degree at Sapienza University of Rome. I have always been interested in applying my technical knowledge to the space sector, particularly in the field of Earth observation.

Current Role

I am currently completing my Master's degree in Artificial Intelligence and Robotics at Sapienza University of Rome. My final thesis will be conducted in collaboration with the ESA Φ-lab, where I will stay as a Visiting Researcher. The research focuses on improving the resilience of deep learning models for Earth Observation against adversarial attacks. The study will investigate methods for detecting adversarial samples, strengthening model robustness, and recovering corrupted satellite data. The work will initially address image classification and may later be extended to semantic segmentation.

Areas of Expertise

My expertise covers several areas of Computer Engineering, including databases, computer architecture, operating systems, software development, and programming. During my Master’s degree, I specialized in Artificial Intelligence, with a particular focus on Machine Learning, Deep Learning, and Neural Networks. Since my Bachelor’s degree, I have also applied these methods to satellite imagery through projects outside the university, gaining practical experience with Earth Observation data and multispectral images. This combination of computer science, AI, and Earth Observation skills enables me to contribute to the ESA Φ-lab CIN by investigating and improving the robustness of deep learning models for satellite data.

Vision for the Future

I envision a future in which Artificial Intelligence becomes an increasingly important tool for extracting reliable and valuable information from Earth Observation data. To achieve this, AI models must not only be accurate but also robust against perturbations, data degradation, and unexpected conditions. My work aims to contribute to this goal by studying how deep learning models can more effectively detect and withstand adversarial attacks while also preserving or recovering valuable satellite data. In the long term, I plan to support the development of more trustworthy AI systems for Earth Observation applications.

Project as CIN Researcher

Coming Soon