CoRReCt: Compute, Record, Replay, Compare to Protect against Hardware Trojans
For the CoRReCt demonstrator, we are applying the approach developed in the research area QuAC to quantify the security of a critical component.
Name | |
---|---|
Beckert, Bernhard | beckert ∂does-not-exist.kit edu |
Dörre, Felix | felix doerre ∂does-not-exist.kit edu |
Mechler, Jeremias | jeremias mechler ∂does-not-exist.kit edu |
Müller-Quade, Jörn | |
Reiche, Frederik | frederik reiche ∂does-not-exist.kit edu |
Reussner, Ralf | ralf reussner ∂does-not-exist.kit edu |
1 additional person visible within KIT only. |
SECAIMED: Secure and Compliant AI for Medical Data
In “SECAIMED”, researchers from DKFZ and Topic ESS jointly develop a novel and legally compliant approach for secure machine learning with applications in medicine.
Secure Computations Using Not-so-Trusted Hardware
Name | |
---|---|
Dörre, Felix | |
Mechler, Jeremias | jeremias mechler ∂does-not-exist.kit edu |
Müller-Quade, Jörn |
Quantitative Analyses of 4Crypt – Privacy-Preserving Documentation for Assembly Assistance Systems
Quantitative Analyses of 4Crypt – Privacy-Preserving Documentation for Assembly Assistance Systems
We present a privacy-friendly and trustworthy interactive assembly table as a showcase for our methodological results with respect to quantification. It uses cameras and AI to provide support during assembly tasks. The video feed can also be recorded for later analysis of critical work steps. While this is useful for quality control, it also entails a major privacy concern, as workers may be subjected to unfounded video surveillance by their employer. We quantify the privacy-related technical mechanisms of 4Crypt with respect to their security as well as their effect on workers.