Non-interleaved chiral metasurfaces for high-spatial-resolution polarimetry are proposed and demonstrated. Furthermore, a convolutional neural network is incorporated to analyze interferometric images with the polarization state of light, and it results in accurate Stokes parameters.

中文翻译：

---

非交错手性超表面和神经网络提高了旋光测定的空间分辨率

提出并论证了用于高空间分辨率偏振测量的非交错手性超表面。此外，还结合了卷积神经网络来分析具有光偏振态的干涉图像，并产生准确的斯托克斯参数。