Presentation Information
[24a-31A-4]Demonstration of the online learning on the magneto-optical diffractive deep neural network
〇(D)Hotaka Sakaguchi1, Riku Oya1, Jian Zhang1, Takuma Honma1, Satoshi Sumi2, Hiroyuki Awano2, Hirofumi Nonaka3, Fatima Zahra Chafi1, Takayuki Ishibashi1 (1.Nagaoka Univ. of Tech., 2.Toyota Tech. Inst., 3.Aichi Inst. of Tech.)
Keywords:
optical computing,magneto-optical effect,neural network
In order to reduce the significant power consumption issue associated with deep learning, we have developed a magneto-optical diffraction deep neural network (MO-D2NN) that operates in the visible light spectrum, utilizing the magneto-optical effects of magnetic materials. MO-D2NN employs rewritable and non-volatile magnetic domains as neurons, enabling online learning, a challenging task for conventional D2NNs. In this presentation, we report the results of foundational experiments aimed at online learning with MO-D2NN.