ONEPLAST Project Goals
The ONEPLAST project aims to develop neuromorphic optical neural networks based on the plasticity of the nonlinear refractive index.
Primary objectives
1
Development of Advanced Optical Memory
Experimental
realization
of episodic memory capable of storing information bit by bit.
A. Bile, H. Tari, E. Fazio, Episodic Memory and Information Recognition Using Solitonic
Neural Networks Based on Photorefractive Plasticity. Appl. Sci. 2022, 12, 5585,
https://doi.org/10.3390/app12115585.
A. Bile, H. Tari, E. Fazio, Episodic Memory and Information Recognition Using Solitonic
Neural Networks Based on Photorefractive Plasticity. Appl. Sci. 2022, 12, 5585,
https://doi.org/10.3390/app12115585.
Designing
an architecture
that implements procedural memory, capable of recognizing and reducing data through keyword extraction
A. Bile, H. Tari, E. Fazio, Episodic Memory and Information Recognition Using Solitonic
Neural Networks Based on Photorefractive Plasticity. Appl. Sci. 2022, 12, 5585,
https://doi.org/10.3390/app12115585.
A. Bile, H. Tari, E. Fazio, Episodic Memory and Information Recognition Using Solitonic
Neural Networks Based on Photorefractive Plasticity. Appl. Sci. 2022, 12, 5585,
https://doi.org/10.3390/app12115585.
Creating
a semantic system
capable of associating labels with previously recognized data.
2
Harnessing Optical Plasticity for Neuromorphic Circuits
Creating
solitonic waveguides
which self-generate and dynamically modify the refractive index in response to stimuli.
A. Bile, M. Chauvet, H. Tari and E. Fazio Supervised Learning of soliton X-junctions in Lithium Niobate films On Insulator, Optics Letters 47, 21 (2022),
https://doi.org/10.1364/OL.468997
A. Bile, M. Chauvet, H. Tari and E. Fazio Supervised Learning of soliton X-junctions in Lithium Niobate films On Insulator, Optics Letters 47, 21 (2022),
https://doi.org/10.1364/OL.468997
Studying
solitonic X-junctions
to develop adaptive interconnections that learn and modify their state based on input signals.
A. Bile, M. Chauvet, H. Tari and E. Fazio Supervised Learning of soliton X-junctions in
Lithium Niobate films On Insulator, Optics Letters 47, 21 (2022),
https://doi.org/10.1364/OL.468997
A. Bile, M. Chauvet, H. Tari and E. Fazio Supervised Learning of soliton X-junctions in Lithium Niobate films On Insulator, Optics Letters 47, 21 (2022),
https://doi.org/10.1364/OL.468997
3
Integration of Optical Neural Networks at the Sub-Micron Scale
Developing
plasmonic circuits
that integrate solitonic technology to achieve ultra-compact, plastic interconnections.
Experimentally demonstrating
the integration
of plasmonic and solitonic signals, enhancing miniaturization and reducing energy consumption.
4
Surpassing Von Neumann Architecture Limitations
Eliminating
the separation
between computation and memory units, adopting a paradigm where information is processed and stored simultaneously.
Demonstrating
that photonic neuromorphic circuits
can partially replace software-based computation, enhancing speed and efficiency.
5
Practical Implementation of Plastic Optical Networks
Eliminating
the separation
between computation and memory units, adopting a paradigm where information is processed and stored simultaneously.
Demonstrating
that photonic neuromorphic circuits
can partially replace software-based computation, enhancing speed and efficiency.
6
Experimental Demonstration on Advanced Substrates
Employing
ultrathin lithium niobate (LNOI) crystals
for the generation of low-power solitons.
Developing
neuromorphic optical devices
with plastic, reconfigurable, and erasable interconnections.