Since ancient times, mapmaking has crossed the border between art and science, exploration and cataloging, adventure and reasoning. Creating digital maps is no different, as the project by computer scientist Jefersson dos Santos shows.
Santos is situated on the flourishing frontier between machine learning and geography. His research seeks to make the processes of algorithms that classify images taken by satellites more robust as a step in creating digital maps. This work is promising for various applications, including monitoring urban housing conditions and their relationship with dengue fever and mapping rural roads.
Santos is passionate about both computer science and geography. He has a bachelor’s degree in computer science from the State University of Mato Grosso do Sul, a master’s degree from the University of Campinas, and a doctorate from the same university, where he also spent time studying at the Université de Cergy-Pontoise in France.
Santos has always focused on interdisciplinarity. He founded and coordinates the Pattern Recognition and Earth Observation Laboratory (PATREO) at the Federal University of Minas Gerais, where he is currently a lecturer.
Santos is from the state of Mato Grosso do Sul and is married to a woman from São Paulo with a PhD in linguistics. They are now the parents of a couple of children born in Minas Gerais. A fan of the Palmeiras football club, Santos defines computing as his vocation and his passion for geography and flags as a hobby.
The creation of geographical maps using satellite images is a supervised classification problem. This means that algorithms are trained to identify patterns of interest from labeled pixels (samples). Deep learning-based techniques have achieved significant advances in this area, but several computational challenges remain, especially when the models are applied to real-world problems with class imbalance, under-representation, and unknown classes. Additionally, pixel annotation is costly and requires expertise in the target application, limiting the volume of available annotated data.
This project aims to tackle these computational challenges to increase the robustness of satellite image classification models. The developed techniques will be evaluated in real-world problems, such as mapping rural roads and monitoring urban housing conditions and their relationship with dengue outbreaks.