Replacing electrons with photons is a compelling route toward high-speed, massively parallel, and low-power artificial intelligence computing. Recently, diffractive networks composed of phase surfaces were trained to perform machine learning tasks through linear optical transformations. However, the existing architectures often comprise bulky components and, most critically, they cannot mimic the human brain for multitasking. Here, we demonstrate a multi-skilled diffractive neural network based on a metasurface device, which can perform on-chip multi-channel sensing and multitasking in the visible. The polarization multiplexing scheme of the subwavelength nanostructures is applied to construct a multi-channel classifier framework for simultaneous recognition of digital and fashionable items. The areal density of the artificial neurons can reach up to 6.25 × 10⁶ mm⁻² multiplied by the number of channels. The metasurface is integrated with the mature complementary metal-oxide semiconductor imaging sensor, providing a chip-scale architecture to process information directly at physical layers for energy-efficient and ultra-fast image processing in machine vision, autonomous driving, and precision medicine.

Diffractive achromats (DAs) promise ultra-thin and light-weight form factors for full-color computational imaging systems. However, designing DAs with the optimal optical transfer function (OTF) distribution suitable for image reconstruction algorithms has been a difficult challenge. Emerging end-to-end optimization paradigms of diffractive optics and processing algorithms have achieved impressive results, but these approaches require immense computational resources and solve non-convex inverse problems with millions of parameters. Here, we propose a learned rotational symmetric DA design using a concentric ring decomposition that reduces the computational complexity and memory requirements by one order of magnitude compared with conventional end-to-end optimization procedures, which simplifies the optimization significantly. With this approach, we realize the joint learning of a DA with an aperture size of 8 mm and an image recovery neural network, i.e., Res-Unet, in an end-to-end manner across the full visible spectrum (429–699 nm). The peak signal-to-noise ratio of the recovered images of our learned DA is 1.3 dB higher than that of DAs designed by conventional sequential approaches. This is because the learned DA exhibits higher amplitudes of the OTF at high frequencies over the full spectrum. We fabricate the learned DA using imprinting lithography. Experiments show that it resolves both fine details and color fidelity of diverse real-world scenes under natural illumination. The proposed design paradigm paves the way for incorporating DAs for thinner, lighter, and more compact full-spectrum imaging systems.

Exceptional points (EPs) can achieve intriguing asymmetric control in non-Hermitian systems due to the degeneracy of eigenstates. Here, we present a general method that extends this specific asymmetric response of EP photonic systems to address any arbitrary fully-polarized light. By rotating the meta-structures at EP, Pancharatnam-Berry (PB) phase can be exclusively encoded on one of the circular polarization-conversion channels. To address any arbitrary wavefront, we superpose the optical signals originating from two orthogonally polarized -yet degenerate- EP eigenmodes. The construction of such orthogonal EP eigenstates pairs is achieved by applying mirror-symmetry to the nanostructure geometry flipping thereby the EP eigenmode handedness from left to right circular polarization. Non-Hermitian reflective PB metasurfaces designed using such EP superposition enable arbitrary, yet unidirectional, vectorial wavefront shaping devices. Our results open new avenues for topological wave control and illustrate the capabilities of topological photonics to distinctively operate on arbitrary polarization-state with enhanced performances.

Optical sorting combines optical tweezers with diverse techniques, including optical spectrum, artificial intelligence (AI) and immunoassay, to endow unprecedented capabilities in particle sorting. In comparison to other methods such as microfluidics, acoustics and electrophoresis, optical sorting offers appreciable advantages in nanoscale precision, high resolution, non-invasiveness, and is becoming increasingly indispensable in fields of biophysics, chemistry, and materials science. This review aims to offer a comprehensive overview of the history, development, and perspectives of various optical sorting techniques, categorised as passive and active sorting methods. To begin, we elucidate the fundamental physics and attributes of both conventional and exotic optical forces. We then explore sorting capabilities of active optical sorting, which fuses optical tweezers with a diversity of techniques, including Raman spectroscopy and machine learning. Afterwards, we reveal the essential roles played by deterministic light fields, configured with lens systems or metasurfaces, in the passive sorting of particles based on their varying sizes and shapes, sorting resolutions and speeds. We conclude with our vision of the most promising and futuristic directions, including AI-facilitated ultrafast and bio-morphology-selective sorting. It can be envisioned that optical sorting will inevitably become a revolutionary tool in scientific research and practical biomedical applications.

Reflecting light to a predetermined nonspecular direction is an important ability of metasurfaces, which is the basis for a wide range of applications (e.g., beam steering/splitting and imaging). However, anomalous reflection with 100% efficiency has not been achieved at optical frequencies yet, because of losses and/or insufficient nonlocal control of light waves. Here, we propose an all-dielectric quasi–three-dimensional subwavelength structure, consisting of multilayer films and metagratings, to achieve perfect anomalous reflections at optical frequencies. A complex multiple scattering process was stimulated by effectively coupling different Bloch waves and propagating waves, thus offering the metasystem the desired nonlocal control on light waves required by perfect anomalous reflections. Two perfect anomalous reflectors were demonstrated to reflect normally incident 1550-nm light to the 40°/75° directions with absolute efficiencies of 99%/99% in design (98%/88% in experiment). Our results pave the way toward realizing optical metadevices with desired high efficiencies in realistic applications.

Accurately and swiftly characterizing the state of polarization (SoP) of complex structured light is crucial in the realms of classical and quantum optics. Conventional strategies for detecting SoP, which typically involves a sequence of cascaded optical elements, are bulky, complex, and run counter to miniaturization and integration. While metasurface-enabled polarimetry has emerged to overcome these limitations, its functionality predominantly remains confined to identifying SoP within the standard Poincaré sphere framework. The comprehensive detection of SoP on the higher-order Poincaré sphere (HOPS), however, continues to be a huge challenge. Here, we propose a general polarization metrology method capable of fully detecting SoP on any HOPS through a single measurement. The underlying mechanism relies on transforming the optical singularities and Stokes parameters into visualized intensity patterns, facilitating the extraction of all parameters that fully determine a SoP. We actualize this concept through a novel meta-device known as the metasurface photonics polarization clock, which offers an intuitive display of SoP using four distinct pointers. As a proof of concept, we theoretically and experimentally demonstrate fully resolving SoPs on the 0th, 1st, and 2nd HOPSs. Our implementation opens up a new pathway towards real-time polarimetry of arbitrary beams featuring miniaturized size, a simple detection process, and a direct readout mechanism, promising significant advancements in fields reliant on polarization.