Quantum theory departs from classical physics in its treatment of correlations, most prominently through the phenomena of contextuality and nonlocality. Once regarded primarily as foundational curiosities, these efects are now understood as key operational resources for quantum computation, communication, and simulation. Although traditionally investigated in distinct settings, recent theoretical and experimental advances have revealed deep conceptual, mathematical, and operational connections between them. This review presents a unified perspective on these developments based on sheaf-theoretic and graph-theoretic frameworks, which provide theory-independent characterizations of statistical correlations. These approaches clarify the structural relationship between contextuality and nonlocality, facilitate the formulation of experimentally testable inequalities, and guide implementations in realistic physical platforms, with particular emphasis on photonic systems. By bridging abstract theoretical structures and concrete experimental realizations, this review sheds light on the nonclassical foundations of quantum correlations and their emerging role in quantum technologies.

Conventional complementary metal-oxide-semiconductor (CMOS) image sensors with Bayer color filter arrays face limitations in low-illumination color imaging. Color router (CR) aims to achieve absorption-loss-free spectral separation, providing an efficient way to sort and guide different color lights to the corresponding pixels. However, applying CR to imaging chips requires consideration of different incidence conditions and processing difficulties. In this study, we propose and fabricate a polarization-insensitive “red, green and blue + near-infrared (NIR)” CR based on a metasurface, operating within the wavelength range of 400-1,100 nm and exhibiting high light energy utilization efficiency. The structural parameters are optimized through an inverse design method. The device is fabricated by silicon-compatible processes, offering the advantage of high integration. Experimental results demonstrate that the measured energy utilization efficiency reaches over 90%, and the measured average correlation coefficient of spectral curves under different polarizations reaches 0.98. This device can efficiently route visible to NIR light, providing a novel approach for integrating metasurfaces into CMOS image sensors to enhance high-performance color imaging systems.

Metasurfaces, as a key platform for flat optics, enable advanced manipulation of light by harnessing rich local and nonlocal resonant modes at subwavelength-structured interfaces. Distinct from conventional local metasurfaces that primarily modulate optical wavefronts over broad spectral ranges, nonlocal metasurfaces based on collective resonances provide enhanced spectral and momentum selectivity together with strengthened light–matter interaction. This review surveys recent advances in optical-field information manipulation enabled by nonlocal metasurfaces, where “information manipulation” denotes the high-dimensional control of amplitude, phase, and polarization across spectral, temporal, spatial, and momentum domains, with a particular focus on integrating local and nonlocal resonances to enhance device performance and functionality. We first introduce the fundamental characteristics of nonlocal resonances and summarize representative progress in real- and momentum-space light control within passive nonlocal metasurfaces. We also discuss recent advances in active nonlocal metasurfaces, including applications in nonlinear harmonic generation, quantum-state control, and spatial information lasers. Finally, we highlight emerging trends in on-chip and tunable nonlocal metasurfaces and outline key challenges and future research directions for this rapidly evolving field.

Phase-shifting profilometry (PSP), as a high-precision and low-cost three-dimensional (3D) profile measurement technique, has been extensively applied in diverse fields. By introducing the compressive sensing paradigm, compressive phase-shifting fringe projection profilometry (CPSFPP) provides an effective solution for high-speed dynamic object acquisition. Nevertheless, achieving high-fidelity dynamic profile reconstruction from compressed measurements remains a considerable challenge, owing to the inherent information loss in compressive sampling and limitations in computational reconstruction, especially under high compression ratios. To address this issue, we propose a physics-informed neural network (PINN)-based computational imaging framework for high-fidelity CPSFPP, named PINN-CPSFPP. This method integrates the physical model constraints with neural network learning to guarantee high-fidelity reconstruction even at high compression ratios. We perform numerical simulations to verify the reconstruction accuracy of PINN-CPSFPP under different compression ratios and experimentally validate the method by measuring translational, rotational, and deformed objects. The results demonstrate that the measurement speed is increased by nine times compared with conventional PSP. Benefiting from its robust 3D imaging performance, PINN-CPSFPP serves as a high-fidelity metrological tool for high-speed 3D scenarios and exhibits promising application prospects in a wide range of basic and applied disciplines.

Metasurfaces composed of subwavelength-scale artificial meta-atoms have emerged as a powerful platform for manipulating light. By enabling strong light-matter interactions within ultrathin planar geometries, metasurfaces have opened new avenues for highly integrated photonic devices. Lithium niobate (LiNbO₃) is considered one of the most promising multifunctional integrated photonic platforms due to its outstanding properties, such as a large second-order nonlinear susceptibility, a broad transparency window, and a strong electro-optic (EO) effect. In recent years, integrated photonic devices based on lithium-niobate-on-insulator platforms have experienced rapid development. This review summarizes recent advances in LiNbO₃ metasurfaces, providing a comprehensive overview of their demonstrated applications in nonlinear frequency conversion, wavefront and phase modulation, and dynamic EO modulation. By systematically introducing the interplay between the intrinsic material properties of LiNbO₃ and the structural design principles of metasurfaces, this review offers a coherent framework for understanding their nonlinear and active optical functionalities, and serves as a valuable reference for the design and implementation of nonlinear and actively tunable micro- and nano-photonic devices.