Reproduce “Federated Neuromorphic Learning of Spiking Neural Networks for Low-Power Edge Intelligence” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Nicolas Skatchkovsky, Hyeryung Jang, Osvaldo Simeone
Researchers at King's College London developed FL-SNN, an online federated learning algorithm for cooperatively training Spiking Neural Networks (SNNs) on edge devices without backpropagation. The framework enhances performance over isolated SNN training on heterogeneous data and achieves better accuracy by exchanging small subsets of weights more frequently.
Key insights
- Combining Spiking Neural Networks with Federated Learning enables energy-efficient, privacy-preserving, and cooperative AI solutions for edge devices.
- The use of local feedback signals (eligibility traces and learning signals) effectively replaces backpropagation for online SNN training within a federated framework.
- Reducing communication load in FL-SNN is more effectively achieved by transmitting a small subset of weights frequently rather than a larger subset infrequently.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.