Enhancing Efficient Continual Learning with Dynamic Structure Development of Spiking Neural Networks

TL;DR

This paper introduces DSD-SNN, a brain-inspired spiking neural network that dynamically adapts its structure for efficient continual learning, improving performance and reducing computational costs.

Contribution

It presents a novel dynamic structure development approach for SNNs, enabling adaptive growth and pruning for continual learning tasks.

Findings

Significantly improves learning speed and memory capacity.

Reduces computational overhead compared to existing methods.

Achieves state-of-the-art results in SNN-based continual learning.

Abstract

Children possess the ability to learn multiple cognitive tasks sequentially, which is a major challenge toward the long-term goal of artificial general intelligence. Existing continual learning frameworks are usually applicable to Deep Neural Networks (DNNs) and lack the exploration on more brain-inspired, energy-efficient Spiking Neural Networks (SNNs). Drawing on continual learning mechanisms during child growth and development, we propose Dynamic Structure Development of Spiking Neural Networks (DSD-SNN) for efficient and adaptive continual learning. When learning a sequence of tasks, the DSD-SNN dynamically assigns and grows new neurons to new tasks and prunes redundant neurons, thereby increasing memory capacity and reducing computational overhead. In addition, the overlapping shared structure helps to quickly leverage all acquired knowledge to new tasks, empowering a single network capable of supporting multiple incremental tasks (without the separate sub-network mask for each task). We validate the effectiveness of the proposed model on multiple class incremental learning and task incremental learning benchmarks. Extensive experiments demonstrated that our model could significantly improve performance, learning speed and memory capacity, and reduce computational overhead. Besides, our DSD-SNN model achieves comparable performance with the DNNs-based methods, and significantly outperforms the state-of-the-art (SOTA) performance for existing SNNs-based continual learning methods.