0

Shrinking Your TimeStep: Towards Low-Latency Neuromorphic Object Recognition with Spiking Neural Network

Neuromorphic object recognition with spiking neural networks (SNNs) is the cornerstone of low-power neuromorphic computing. However, existing SNNs suffer from significant latency, utilizing 10 to 40 timesteps or more, to recognize neuromorphic objects.

Year
2024
Hosting
External sourcelicense unknown

Cite

Notes

Only stored in your browser.

Attribution

Abstract & full text
arxiv.org/abs/2401.01912v1
TL;DR
Semantic Scholar
Attribution policy →