Boosting the Robustness-Accuracy Trade-off of SNNs by Robust Temporal Self-Ensemble
This work addresses adversarial robustness in SNNs, which is crucial for energy-efficient and brain-inspired computing, but it appears incremental as it builds on temporal ensembling concepts.
The paper tackled the vulnerability of Spiking Neural Networks (SNNs) to adversarial perturbations by proposing Robust Temporal self-Ensemble (RTE), a training framework that improves robustness and reduces temporal transferability of attacks, resulting in consistent outperformance of existing methods in robust-accuracy trade-off across multiple benchmarks.
Spiking Neural Networks (SNNs) offer a promising direction for energy-efficient and brain-inspired computing, yet their vulnerability to adversarial perturbations remains poorly understood. In this work, we revisit the adversarial robustness of SNNs through the lens of temporal ensembling, treating the network as a collection of evolving sub-networks across discrete timesteps. This formulation uncovers two critical but underexplored challenges-the fragility of individual temporal sub-networks and the tendency for adversarial vulnerabilities to transfer across time. To overcome these limitations, we propose Robust Temporal self-Ensemble (RTE), a training framework that improves the robustness of each sub-network while reducing the temporal transferability of adversarial perturbations. RTE integrates both objectives into a unified loss and employs a stochastic sampling strategy for efficient optimization. Extensive experiments across multiple benchmarks demonstrate that RTE consistently outperforms existing training methods in robust-accuracy trade-off. Additional analyses reveal that RTE reshapes the internal robustness landscape of SNNs, leading to more resilient and temporally diversified decision boundaries. Our study highlights the importance of temporal structure in adversarial learning and offers a principled foundation for building robust spiking models.