I am a PhD student at Doctoral School at TU Wien Informatics advised by Andreas Rauber. In my dissertation, I focus on novel approaches for ownership verification of tabular data that preserve multiple aspects of data utility. Currently, I am a researcher in SBA Research in the Machine Learning and Data Management Group group where I work on projects in the areas of data privacy, ownership verification and ML privacy & security.
Currently, I am investigating ethical approaches to generative AI, focusing on developing trustworthy systems that enhance rather than disrupt artistic and human creativity, thereby fostering innovation through responsible implementation and use.
I previously completed master studies Logic and Computation at TU Wien, and the bachelor program in Computer Science at the Faculty of Electrical Engineering and Computing at Zagreb University. I have been tutoring and teaching at TU Wien and FH Technikum in courses Machine Learning and Security and Privacy in Machine Learning.
Curriculum Vitae (last update: November 25, 2024)
✨ I am passionate about Machine Learning and, in particular, how to make it secure, safe and privacy-aware. ✨
Publications
All publications are listed on my Google Scholar account. The following list is last updated on August 7, 2024:
- U Can’t Gen This? A Survey of Intellectual Property Protection Methods for Data in Generative AI.
Šarčević, T., Karlowicz, A., Mayer, R., Baeza-Yates, R. and Rauber, A.
arXiv preprint arXiv:2406.15386 (2024)
[code] - Achieving Privacy and Tracing Unauthorised Usage: Anonymisation-based Fingerprinting of Private Data.
Šarčević, T., Mayer, R. and Adler, P., 2023, December.
In 2023 IEEE International Conference on Big Data (BigData) (pp. 5578-5587). IEEE. - Adaptive attacks and targeted fingerprinting of relational data.
Šarčević, T., Mayer, R. and Rauber, A., 2022, December.
In 2022 IEEE International Conference on Big Data (Big Data) (pp. 5792-5801). IEEE. - Semantic-enabled architecture for auditable privacy-preserving data analysis.
Ekaputra, F.J., Ekelhart, A., Mayer, R., Miksa, T., Šarčević, T., Tsepelakis, S. and Waltersdorfer, L., 2021.
Semantic Web, pp.1-34. - An Analysis of Different Notions of Effectiveness in k-Anonymity.
Šarčević, T., Molnar, D. and Mayer, R., 2020, September.
In International Conference on Privacy in Statistical Databases (pp. 121-135). Springer, Cham. - A Correlation-Preserving Fingerprinting Technique for Categorical Data in Relational Databases.
Šarčević, T. and Mayer, R., 2020, September.
In IFIP International Conference on ICT Systems Security and Privacy Protection (pp. 401-415). Springer, Cham.
[code] - An Evaluation on Robustness and Utility of Fingerprinting Schemes.
Šarčević, T. and Mayer, R., 2019, August.
In International Cross-Domain Conference for Machine Learning and Knowledge Extraction (pp. 209-228). Springer, Cham. - Artificial Bee Colony Algorithm for Solving the Flight Disruption Problem.
Šarčević, T., Rocha, A.P. and Castro, A.J., 2018, June.
In International Conference on Practical Applications of Agents and Multi-Agent Systems (pp. 72-81). Springer, Cham.
Project highlights
Here are listed some of my recent projects:
| Project | About | Source/Publications |
|---|---|---|
| GenArt | Art in the age of generative AI: exploring sustainable coexistence of human creations and AI-generated content. | paper repo |
| NCorr-FP | Fingerprinting for tracing unauthorised usage of data. | source |
| Monitaur | Monitoring solution for detecting malicious interaction with ML models online. | |
| IPP4ML | Intellectual property protection in ML processes: protecting training data and measuring impact on ML models. | toolbox CD-MAKE 2019 IFIP SEC 2020 IEEE BigData 2022 IEEE BigData 2023 |
| Fed-WM | Watermarking federated ML models | poster |
| WellFort | A platform for privacy-preserving data analysis | ERCIM NEWS 2021 Semantic Web Journal |
| BeyondCoding | Coaching programme for efficient, agile and secure software development. |