Trickster
Sophisticated attacks on Machine Learning models
Trickster allows to attack machine learning models where the attack needs to be more sophisticated than simply adding some noise to an image. One example is to evade an abuse detection model, for example, social media bot detection, or a malware detector.
Attack
- Technical
- Research papers
- Source code: Lab GitHub
- Last commit: 2019-04-18
Security and Privacy Engineering Laboratory
Security and Privacy Engineering Laboratory
The Security and Privacy Engineering Laboratory develops tools and methodologies to help engineers building systems that respect societal values, such as security, privacy or non discrimination. Currently, they are working on
- Machine Learning impact on society
- Evaluating privacy in complex systems
- Engineering privacy-preserving systems
This page was last edited on 2022-07-07.
This page was last edited on 2022-07-07.