Perceptrons with Hebbian learning based on wave ensembles in plastic potentials

Abstract

A general scheme to realize a perceptron for hardware neural networks is presented, where multiple interconnections are achieved by a superposition of Schrodinger waves. Spatially patterned potentials process information by coupling different points of reciprocal space. The necessary potential shape is obtained from the Hebbian learning rule, either through exact calculation or construction from a superposition of known optical inputs. This allows implementation in a wide range of compact optical systems, including: 1) any non-linear optical system; 2) optical systems patterned by optical lithography; and 3) exciton-polariton systems with phonon or nuclear spin interactions.

References in corpus (7)

• Exciton-polariton condensates
• Dynamics of Quantum Dot Nuclear Spin Polarization Controlled by a Single Electron
• Light Engineering of the Polariton Landscape in Semiconductor Microcavities
• Bistability of the Nuclear Polarisation created through optical pumping in InGaAs Quantum Dots
• Local tuning of photonic crystal cavities using chalcogenide glasses
• Subwavelength modulational instability and plasmon oscillons in nanoparticle arrays
• Optically erasing disorder in semiconductor microcavities with dynamic nuclear polarization