Luca Bortolussi

Blending adversarial training and representation-conditional purification via aggregation improves adversarial robustness

2025 · ArTS Archivio della ricerca di Trieste (University of Trieste https://www.units.it/)

In this work, we propose a novel adversarial defence mechanism for image lassification - CARSO - blending the paradigms of adversarial training and adversarial purification in a synergistic robustness-enhancing way. The method builds upon an …