Spiffy: Efficient Implementation of CoLaNET for Raspberry Pi
This provides a lightweight software solution for deploying SNNs on resource-constrained devices, but it is incremental as it applies an existing method to new hardware.
The paper tackled running spiking neural networks efficiently on common hardware by implementing CoLaNET in Rust and optimizing it for platforms like Raspberry Pi, achieving 92% accuracy on MNIST with 0.9 ms per training step and 0.45 ms per inference step.
This paper presents a lightweight software-based approach for running spiking neural networks (SNNs) without relying on specialized neuromorphic hardware or frameworks. Instead, we implement a specific SNN architecture (CoLaNET) in Rust and optimize it for common computing platforms. As a case study, we demonstrate our implementation, called Spiffy, on a Raspberry Pi using the MNIST dataset. Spiffy achieves 92% accuracy with low latency - just 0.9 ms per training step and 0.45 ms per inference step. The code is open-source.