The first optical neural network is
Trained multilayer phase mask (classifier of handwritten symbols). On the right is the physical model of the D²NN optical neural network, printed on a 3D printer: 8 × 8 cm layers with a distance of 3 cm between each other
A group of researchers from the University of California, Los Angeles, developed a new type of neural network that uses light instead of electricity in its work. In the journal Science published article with the description of the idea, the working device, its performance and types of applications, which, according to their authors, can be well calculated in neural networks of a new type.
Fully optical Diffractive Deep Neural Network (D²NN), which is physically formed from a variety of reflective or transparent surfaces. These surfaces work together, performing an arbitrary function, acquired as a result of training. While obtaining the result and forecasting in the physical network is organized completely optically, the training part with the design of the reflective surface structure is calculated on the computer.
published July 2? 2018 in the journal Science (doi: ??? /science.aat808? .pdf ).
It may be interesting
5f221f0ae514f1ec90f2071d877f5494#p713382 https://seomarketer.gr/2020/07/anakainisi-xenodoxeiou-pote-einai-aparaitito/ https://seomarketergr.com/4-tropoi-gia-na-veltiosete-to-xenodoxeio-sas-logo-cov
id-19/ https://gamereviews.gr/5-logoi-gia-anakainisi-ksenodoxeiou/ http://jifficlassified.ca/index.php?page=user&action=pub_profile&id=180
170 http://proothisistoselidonseogr.over-blog.com/2020/07/.html https://ko-fi.com/post/Βασικα-στοιχεια-λαμπτηρα-Wattage-και-Lumens-Q5Q51WG5E http://www.true2ourselves.com/blogs/blog/index.php?seid=georgetech&id=29955 https://www.openlearning.com/u/innovation-qdgs3n/