Abstract: Spiking Neural Networks (SNNs) have been in-creasingly investigated for audio recognition due to the low power consumption on neuromorphic hardware by mimicking biological neural systems.