We provide a protocol for training photonic neural networks based on adjoint methods. The gradient of the network with respect to its tunable degrees of freedom is computed by physically backpropagating an optical error signal. © 2019 The Author(s)
We provide a protocol for training photonic neural networks based on adjoint methods. The gradient of the network with respect to its tunable degrees of freedom is computed by physically backpropagating an optical error signal. © 2019 The Author(s)