UCLA researchers create all-optical diffractive deep neural network (2018)
photonics.com
photonics.com
The non-lineartiy (or activation function) is crucial to the high performance we see with deep neural networks. For example, on the task of interest here (MNIST, a classic hand-written digit classification task), a old-school linear classifier can get between 88% and 92.4% accuracy [1]. Whereas non-linear neural networks easily achieve 95%+ [1].
So until they figure out a way to represent a nice non-linearity (unless I missed something), these all optical techniques are at a permanant disadvantage.
https://advances.sciencemag.org/content/5/12/eaay6946
And the repo: https://github.com/fancompute/wavetorch
https://scholar.google.com/scholar?cites=6457098903910716621...