Camila Roa

Camila Roa
Camila Roa uses self-supervised learning to build representation spaces for retinal image datasets, and then visualises them in 2D using neighbor embedding algorithms. She is a PhD student and part of the IMPRS-IS graduate school. Her research interests include interpretability and robustness in medical applications of machine learning.
Interpretability of ML models is a critical aspect when designing a system to support clinical decision-making.

Present Positions And Title

PhD Student

Social Media

Career

PeriodInstitutionRole
Since 2025University of TübingenPhD student
2023 - 2024Oak Ridge National Laboratory, Center for AI Security ResearchGraduate Research Assistant
2023University of Southern California, Information Sciences InstituteSummer Intern
2022 - 2023University of Tennessee, KnoxvilleGraduate Research Assitant

Academic Education

YearDegreeInstitutionField of Study
2024MScUniversity of Tennessee, KnoxvilleComputer Science
2022BScPontificia Universidad JaverianaElectronics Engineering