Efficient Parallel Training Methods for Spiking Neural Networks with Constant Time Complexity
This addresses computational inefficiency for researchers and practitioners using SNNs, particularly in long-term tasks, though it appears incremental as it builds on existing parallel spiking neuron methods.
They tackled the high time complexity of training Spiking Neural Networks (SNNs) by proposing a Fixed-point Parallel Training (FPT) method, which reduces time complexity from O(T) to O(K) with K as a small constant (e.g., 3) while maintaining accuracy.
Spiking Neural Networks (SNNs) often suffer from high time complexity $O(T)$ due to the sequential processing of $T$ spikes, making training computationally expensive. In this paper, we propose a novel Fixed-point Parallel Training (FPT) method to accelerate SNN training without modifying the network architecture or introducing additional assumptions. FPT reduces the time complexity to $O(K)$, where $K$ is a small constant (usually $K=3$), by using a fixed-point iteration form of Leaky Integrate-and-Fire (LIF) neurons for all $T$ timesteps. We provide a theoretical convergence analysis of FPT and demonstrate that existing parallel spiking neurons can be viewed as special cases of our proposed method. Experimental results show that FPT effectively simulates the dynamics of original LIF neurons, significantly reducing computational time without sacrificing accuracy. This makes FPT a scalable and efficient solution for real-world applications, particularly for long-term tasks. Our code will be released at \href{https://github.com/WanjinVon/FPT}{\texttt{https://github.com/WanjinVon/FPT}}.