Nonlinear photonic lanterns for spiking neural networks
Description: Spiking neural networks (SNN) are bioinspired computing paradigms that enable machine learning with sparse and limited training data. In a novel approach, we investigate photonic platforms that would permit the realization of high optical SNNs, leading to parallelized, ultrafast and energy-efficient operation. In this project, we propose the realization of an important and novel building block of optical SNNs – A optical lantern with engineered coupling between its cores. This architecture, reminiscent of a multimode interferometer, when combined with optical nonlinearities potentially adds nonlinear operations to the optical computing toolkit. This opens new avenues in fundamental research for artificial intelligence and explore the potential of optical systems to build highly energy efficient computing devices.
Funder: Agence Nationale de la Recherche
Amount: €298,600
Funding type: research
Funder scheme: AAPG2023
Start date: 2023-10-01
Investigator: Siddharth SIVANKUTTY
Institution: UMR 8523 - PHLAM - Physique des lasers, atomes et molécules
Funded outputs count: 2
Funder award ID: ANR-23-CE24-0003
OpenAlex ID: G3429720