Development of Fabrication Techniques for Magneto-Optical Diffractive Deep Neural Networks
Hotaka Sakaguchi, Takumi Fujita, Jian Zhang, Satoshi Sumi, Hiroyuki Awano, Hirofumi Nonaka, Takayuki Ishibashi · IEEE Transactions on Magnetics · 2023
Magneto-optical diffractive deep neural network (MO-D2NN) is a type of diffractive deep neural network that utilizes the magneto-optical (MO) effect of magnetic materials. MO-D2NN is composed of magnetic material, which is non-volatile and allows neurons to be rewritten. In our work, we fabricated and evaluated MO-D2NN with two layers containing$100\times100$magnetic domains with a domain size of 1$\mu \text{m}$. Nd0.5Bi2.5Fe4GaO12 thin films were prepared on both sides of a Gd3Ga5O12 substrate and magnetic domain patterns were recorded using the MO recording technique. The recording accuracies of the first and second layers were 79% and 70%, respectively. We believed that MO recording technology can achieve a recording accuracy of over 90% by optimizing the fabrication and recording conditions. The optical setup was built and handwritten digit classification was carried out using fabricated MO-D2NN, but the number of classifications was lower than in the case of simulation. We considered that recording errors were responsible for the low number of classifications and assessed the effect of recording errors on accuracy by simulation. When the recording error was lower than 10%, the loss in classification accuracy was found to be only ~3%. Therefore, we are convinced that MO-D2NN has high physical implementation potential.