Recently, there has been tremendous research studies in optical neural networks that could complete comparatively complex\ncomputation by optical characteristic with much more fewer dissipation than electrical networks. Existed neural networks based\non the optical circuit are structured with an optical grating platform with different diffractive phases at different diffractive points\n(Chen and Zhu, 2019 and Mo et al., 2018). In this study, it proposed a multiwave deep diffractive network with approximately 106\nsynapses, and it is easy to make hardware implementation of neuromorphic networks. In the optical architecture, it can utilize\noptical diffractive characteristic and different wavelengths to perform different tasks. Different wavelengths and different tasks\ninputs are independent of each other. Moreover, we can utilize the characteristic of them to inference several tasks, simultaneously.\nThe results of experiments were demonstrated that the network could get a comparable performance to singlewavelength\nsingle-task. Compared to the multinetwork, single network can save the cost of fabrication with lithography. We train\nthe network on MNISTand MNIST-FASHION which are two different datasets to perform classification...............
Loading....