Retreat, January 26 and 27 2021
The first retreat of the Private AI Institute was a great success: More than 80 participants, lively talks and discussions in which participating PIs and industry partners developed a mutual understanding of the collaborative research, explored opportunities for scientific as well as technical collaboration and discussed future directions.
Sessions and Talks
Privacy of Federated Machine Learning
Security of Machine Learning
Cryptography and Multi-Party Computation
Trusted Execution and Hardware Acceleration
Open Source, Open Data, and Applications
Privacy of Federated Machine Learning
- Federated Learning (survey)
- Robust Knowledge Transfer for Federated Learning
- Group Knowledge Transfer: Federated Learning of Large convolutional neural networks (CNNs) at the Edge
- Federated Multi-Tasking-Learning
Security of Machine Learning
- A Taxonomy of Attacks on Federated Learning
- Intellectual Property (IP) Protection / Model Stealing
- Poisoning Defences for Federated Learning: Goals, Challenges and Solution Approaches
Cryptography and Multi-Party Computation
- MP2ML: A Mixed-Protocol Machine Learning Framework for Private Inference
- A Scalable Approach for Privacy-Preserving Collaborative Machine Learning
Trusted Execution and Hardware Acceleration
- State of the Art of TEE Architectures and Applications to Machine Learning
- Machine Learning on Encrypted Data: Hardware-Software Codesign
- Role of Trusted Execution Environments in PPML
Open Source, Open Data, and Applications
- Open Source Frameworks and Plans for Federated Machine Learning
- Machine Learning applied to malware detection/classification and its extent to Federated Learning
- FedML: A Research Library and Benchmark for Federated Machine Learning
- Cyber-Risk Intelligence Sharing using Federated Learning