This research area focuses on the analysis of security and privacy issues in multimedia and biometric systems, as well as the study of inference risks associated with data-driven and generative models.
It also explores multimodal approaches, such as vision-language models, to enhance forensic analysis and enabling data augmentation in challenging or constrained scenarios.
Key research topics include:
- Biometric authentication and user identification
- Privacy and security in biometric systems
- Information leakage and inference attacks
- Dataset inference in generative models
- Multimodal approaches for forensic analysis and data augmentation
- Multimedia-based information extraction and analysis
- Fairness and robustness in machine learning systems
Selected publications:
Hand Me Your PIN! Inferring ATM PINs of Users Typing with a Covered Hand Proceedings Article
In: 31st USENIX Security Symposium (USENIX Security 22), pp. 1687–1704, 2022.
A study on the impact of multiview distributed feature coding on a multicamera vehicle tracking system at roundabouts Journal Article
In: IEEE Access, vol. 10, pp. 39502–39517, 2022.
Real or Virtual: A Video Conferencing Background Manipulation-Detection System Journal Article
In: arXiv preprint arXiv:2204.11853, 2022.
Responsible innovation at work: gamification, public engagement, and privacy by design Journal Article
In: Journal of Responsible Innovation, vol. 9, no. 3, pp. 315–343, 2022.
Do Not Deceive Your Employer with a Virtual Background: A Video Conferencing Manipulation-Detection System Journal Article
In: arXiv preprint arXiv:2106.15130, 2021.