Summary

2023

Session Number:C3L-4

Session:

Number:C3L-42

GHz-Rate Photonic Spiking Neurons and Spiking Neural Networks

Owen-Newns Dafydd,  Zhang Weikang,  Hejda Matej,  Robertson Joshua,  Hurtado Antonio,  

pp.577-580

Publication Date:2023-09-21

Online ISSN:2188-5079

DOI:10.34385/proc.76.C3L-42

PDF download (925.7KB)

Summary:
Photonic approaches emulating the powerful computa-tional capabilities of the brain are receiving increasing re-search interest for radically new paradigms in ultrafast neuromorphic (brain-like) information processing and Ar-tificial Intelligence (AI). In this talk, I will review our re-search on neuromorphic photonic systems using artifi-cial optical spiking neurons and spiking neural networks for computation. I will describe their potentials for high-speed (GHz rates), low energy photonic spike processing systems for use in strategic applications (e.g. pattern recognition, image processing, data classification). During the talk, I will also introduce our work on laser-based, Re-current Neural Networks for photonic Reservoir Compu-ting systems, yielding excellent performance across com-plex computing tasks at ultrafast rates.