Spectropolarimetric Holographic Multiplexing Metasurface With Super-High Capacity Empowered by Mechanical Rotation.

Mechanically reconfigurable metasurfaces capable of translation, rotation, and permutation have garnered significant interest for high-capacity optical information storage and full-color holographic displays, because of their exceptional functional scalability. Here, we introduce a differentiable inverse design framework for such metasurfaces, establishing an accurate mapping between meta-atom geometries and their multichannel optical responses across diverse optical dimensions, including color and polarization. Leveraging a deep neural network-driven, end-to-end optimization pipeline, the framework enables intelligent, iterative refinement of rotatable metasurfaces within a constrained design space. Using this approach, we demonstrate a high-quality holographic video display by rotating a single element in a cascaded metasurface doublet around the optical axis. The doublet facilitates pixel-resolved holographic imaging with 288 independent channels, and by switching input/output polarization states, the system could support four distinct full-color dynamic holographic videos. This work establishes a novel paradigm for optical parameter multiplexing and end-to-end inverse design in mechanically reconfigurable metasurfaces, with promising applications in compact optical systems, dynamic holography, information processing, and optical computing.