Journal Article10.1063/pt.vbbo.lurd
Nonlinear optical computing doesn’t need nonlinear optics
Johanna L. Miller
1
TL;DR: Researchers propose a novel approach to nonlinear optical computing, bypassing the need for nonlinear optics by encoding data in a way that enables light-based neural networks to function without requiring nonlinear optical components.
read more
Abstract: A major stumbling block on the road to light-based neural networks can be overcome by flipping the script on how data are encoded.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
References
Nonlinear optical encoding enabled by recurrent linear scattering
Fei Xia,Kyung-Tack Kim,Yaniv Eliezer,SeungYun Han,Liam Shaughnessy,Sylvain Gigan,Hui Cao +6 more
TL;DR: Researchers introduce a design leveraging multiple-scattering cavities to induce nonlinear optical mapping, enabling efficient optical data compression and processing for applications like classification, image reconstruction, and object detection with high performance at extreme compression ratios.
10
Tunable nonlinear optical mapping in a multiple-scattering cavity.
TL;DR: Tunable nonlinear optical mapping in a multiple-scattering cavity achieves high-order nonlinear mapping at low optical power by reconfiguring the optical scattering potential. The order of nonlinear mapping can be tuned over a wide range, and the method is robust and free of temporal instability.
10
Nonlinear Processing with Linear Optics
Mustafa Yildirim,Niyazi Ulas Dinc,Ilker Oguz,Demetri Psaltis,Christophe Moser +4 more
- 17 Jul 2023
TL;DR: In this paper , the authors present a framework that uses multiple scattering that is capable of synthesizing programmable linear and nonlinear transformations concurrently at low optical power by leveraging the nonlinear relationship between the scattering potential, represented by data, and the scattered field.
Fully nonlinear neuromorphic computing with linear wave scattering
Clara C. Wanjura,Florian Marquardt +1 more
TL;DR: Researchers propose a neuromorphic computing approach using linear wave scattering, achieving nonlinear processing with high expressivity, and demonstrate its feasibility with integrated photonics, offering a scalable and energy-efficient solution for deep learning applications.
9
Deep learning opens up protein science’s next frontiers
TL;DR: Computer models can now provide stunningly accurate predictions of proteins’ three-dimensional structures, but what about their biological functions?
8