This is a linear setup, thus only ~ 90% on mnist max. This problem is not even discussed in the paper and there are good reasons for that. Nonlinearity in optics requiers either extremely high intensities or 'slow' medium (which would kill the 'speed of light' idea).
All-optical machine learning using diffractive deep neural networks: http://science.sciencemag.org/content/early/2018/07/25/science.aat8084 …https://twitter.com/fchollet/status/958809646885617664 …
-
-
-
The “Speed of light” claim could indeed suffer, but would it make sense energy wise overall?
- Show replies
New conversation -
-
-
I remember seeing similar stuff in EDN back in the 80s - plus ca change ...
-
Here's a typical example from 1987, and I'm sure that wasn't the firsthttps://www.osapublishing.org/ao/abstract.cfm?uri=ao-26-23-5061 …
End of conversation
New conversation -
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.