Metasurface-Based\nIntelligent Identification of Total\nAngular Momentum Spectra for Beams

The total angular momentum (TAM), consisting of spin\nangular momentum\n(SAM) and orbital angular momentum (OAM), is a crucial indicator for\ncharacterizing the topological features of structured beams. However,\ncurrent diagnostic methods have limited measurable modes, making it\ndifficult to obtain the TAM spectrum. Here, we present a metasurface-based\nintelligent scheme for measuring the TAM spectrum. We designed and\nfabricated a metasurface to transform the TAM modes into Hermite–Gaussian-like\nmodes for simplifying judgment and developed a deep learning network,\nwhose core stages are several mobile inverted bottleneck convolution\nlayers for mode decomposition, for accurate TAM spectrum identification.\nThe favorable experimental results demonstrate that our proposal can\nprecisely measure structured beams carrying up to 34 TAM modes. Furthermore,\nrobustness tests of this proposal under noise, angular shift, and\ntransverse rotation demonstrate that our model is capable of accurate\nperformance in the presence of these adverse effects within a certain\nrange. This work presents a new path for measuring the TAM spectrum\nin a miniaturized form, with high accuracy, simple operation, and\nwide measurable modes range, which will inspire more cutting-edge\nscenarios such as laser communication, high security holographic encryption,\nand quantum information processing.