🎯 Our Objectives
TruBrain is a cutting-edge research project focused on building secure, privacy-preserving, and energy-efficient artificial intelligence by combining spiking neural networks (SNNs), federated learning, and neuromorphic hardware.
This project is supported by CHIST-ERA and brings together leading researchers from Queen’s University Belfast, Sorbonne Université, EPFL, and TÜBİTAK.
TruBrain investigates how neuromorphic systems can enhance the trustworthiness of distributed AI by:
🛡️ Enhancing Security: Analyzing vulnerabilities in neuromorphic hardware and federated learning (FL) architectures.
🔒 Preserving Privacy: Studying how SNNs naturally resist privacy attacks like membership inference.
🧩 Bridging Theory and Practice: Translating research into secure and scalable AI implementations.
💻 Building Hardware: Prototyping an FPGA-based federated learning platform with RISC-V and SNNs.