Vmem-$\varphi$: Low-Compute Out-of-Distribution Detection in Spiking Neural Networks from Membrane-Potential Statistics

2026-09-29

AI Summary

This study addresses the challenge of out-of-distribution (OOD) detection in spiking neural networks (SNNs), which typically requires substantial computational resources or access to unavailable output layers. To overcome this, we pioneer the use of subthreshold membrane potential dynamics as a low-cost signal for OOD detection. Methodologically, we propose a corruption-blind multi-descriptor deviation algorithm and construct Gen1-C, an event camera-based data augmentation benchmark. Experimental results demonstrate that under maximum severity levels, the AUROC exceeds 0.88 for five out of six corruption types. These findings effectively validate the feasibility and value of leveraging membrane potentials as a computationally efficient signal for OOD detection in SNNs.

Abstract

Spiking Neural Networks (SNNs) offer an energy-efficient approach to processing event-camera data, yet out-of-distribution (OOD) detection remains challenging in this setting. Existing OOD detection methods often depend on model outputs or computational components that are unavailable in object detection SNNs or are poorly suited to low-compute deployment. To that effect, we show that the subthreshold membrane potential \(V_{\mathrm{mem}}(t)\) provides a useful internal signal for detecting distribution shifts. Simple per-channel statistics derived from these membrane dynamics enable OOD detection. To evaluate this approach, we introduce Gen1-C, an event-camera corruption benchmark developed upon the Prophesee Gen1 automotive detection dataset, containing six sensor-motivated histogram-level stress tests at five severity levels. We further propose the Multi-Descriptor Deviation (MDD), a corruption-blind method that operates on membrane-potential statistics. At the highest corruption severity, MDD achieves an AUROC of more than 0.88 on five of the six corruptions using only a bounded 64-frame observation window. Notably, the remaining corruption is also the one that has the smallest effect on the underlying detector. These results show that the temporal membrane-potential dynamics can provide an effective and low-cost signal for OOD detection in SNN-based event perception.

Problem

• Out-of-Distribution Detection
• Spiking Neural Networks
• Event Cameras
• Low-Compute
• Membrane Potential

Innovation

• Spiking Neural Networks
• Out-of-Distribution Detection
• Membrane Potential
• Event Camera
• Multi-Descriptor Deviation