Methods and systems for designing image processing devices

A system for designing an image processing device comprising at least one metasurface having one or more physical features and a neural network having one or more architecture features configured to perform image reconstruction, comprising:

• one or more processors;
• and
• computer storage storing executable computer instructions in which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
• defining a loss function within a simulation space composed of a plurality of voxels, the simulation space encompassing the one or more features of the metasurface and the one or more architecture features of the neural network, the loss function corresponding to at least one image quality metric resulting from:
• an interaction between an input electromagnetic field at an operative wavelength of the image processing device and the one or more physical features of the metasurface;
• and
• processing, by the neural network, the result of the interaction between the input electromagnetic field and the one or more features of the metasurface;
• defining an initial structure for the one or more physical features of the metasurface and the one or more architectural features of the neural network in the simulation space, at least some of the voxels corresponding to each of the one or more physical features and having a dimension smaller than the operative wavelength;
• determining, using a computer system, values for at least one structural parameter, and/or at least one functional parameter for the one or more physical features and at least one architectural parameter for the one or more architectural features, using a numerical solver to solve Maxwell's equations so that a loss determined according to the loss function is within a threshold loss, wherein the at least one structural and/or functional parameter correspond to a structure of the metasurface and wherein the at least one architectural parameter corresponds to a structure of the neural network;
• defining a final structure of the metasurface based on the values for the one or more structural parameters;
• and
• defining a final structure of the neural network based on the values for the at least one architectural parameter.