Cross-species pose estimation plays a vital role in studying neural mechanisms and behavioral patterns while serving as a fundamental tool for behavior monitoring and prediction. However, conventional image-based approaches face substantial limitations, including excessive storage requirements, high transmission bandwidth demands, and massive computational costs. To address these challenges, we introduce an image-free pose estimation framework based on single-pixel cameras operating at ultra-low sampling rates (6.260 × 10-4). Our method eliminates the need for explicit or implicit image reconstruction, instead directly extracting pose information from highly compressed single-pixel measurements. It dramatically reduces data storage and transmission requirements while maintaining accuracy comparable to traditional image-based methods. Our solution provides a practical approach for real-world applications where bandwidths and computational resources are constrained.

Polymer-embedded liquid crystals (LCs) play a pivotal role in smart applications, enabling precise tunability over electro-optical properties. However, high energy consumption in conventional LC-polymeric systems limits their efficiency in sustainable and environmental protection technologies. Reducing driving voltage without compromising mechanical and electro-optical performance remains an unresolved challenge. Here, we demonstrate a polymer-confined ferroelectric nematic (NF) liquid crystal system, polymerized with mesogenic and non-mesogenic monomers under an electric field. The effective multidomain polymer structure exploits the intriguing properties of the NF LC and generates a highly scattered state with an excellent contrast ratio in the NF phase. Electric field-controlled reorientation of directors leads to a transparent state at a very small voltage. The system demonstrates the advantages of a low driving voltage, sub-millisecond switching time with negligible hysteresis, and improved durability, promoting applications in energy-saving smart windows. This work reveals valuable insights into leveraging NF LCs and tailoring polymer networks to advance the performance of electro-optic devices.

Silicon-based photodetectors are experiencing significant demand for realizing infrared photodetection, night vision imaging, and ultraviolet-enhanced monitoring and communication. Recently, femtosecond-laser (fs-laser) hyperdoped silicon photodetectors have gained attention as promising alternatives to conventional silicon-based devices, owing to their exceptional properties, including high detectivity at low operating bias, broadband response spectrum beyond the bandgap limitation, wide operational temperature range, and ultrahigh dynamic range. Despite these advantages, the practical application of fs-laser hyperdoped devices has been hindered by challenges such as uneven surface structures and numerous lattice defects, which impede industrialization, chip integration, and ultraviolet photodetection performance. In this study, we present, to our knowledge, a novel design of flat fs-laser hyperdoped silicon materials and photodetectors tailored for complementary metal-oxide-semiconductor (CMOS) compatibility. A key innovation lies in the reduction of surface structure dimensions by three orders of magnitude, enabling the integration of fs-laser hyperdoped silicon as a photodetection layer in back-illuminated CMOS devices. The proposed photodetector achieves a peak responsivity of 120.07 A/W and a specific detectivity of 1.27 × 1014 Jones at 840 nm, marking the highest performance reported for fs-laser hyperdoped silicon photodetectors. Furthermore, it demonstrates ultraviolet enhancement and sub-bandgap infrared photodetection simultaneously, with responsivities exceeding 10 A/W across a broad spectrum from 350 to 1170 nm at 5 V. This breakthrough not only paves the way for fs-laser hyperdoped silicon in array photodetection but also facilitates its integration with silicon-based chip fabrication processes, addressing critical bottlenecks for industrialization and advancing the field of silicon photonics.

In contrast to traditional physical measurement methods, machine-learning-based precision measurement is a “data-driven” approach that constitutes a new field of research. We report a machine-learning-based precision measurement of a rotational angle from a vortex-mode shear interferometer, as the two-dimensional optical images at different angles contain the interference patterns that are inherently encoded into the light orbital angular momentum states. Through our evaluation of different convolutional neural networks, we have determined that the ResNeXt50 model excels in detecting minute angle changes across resolutions of 0.05°, 0.1°, 0.5°, 4°, and 10°. This model for the vortex beams achieves over 99.9% accuracy for resolutions of 0.1°, 0.5°, 4°, and 10°, and over 97.0% accuracy for the highest 0.05° resolution. The new results in experiments and modeling demonstrate a robust, accurate, and scalable approach to high-precision rotational angle measurement.

We demonstrate a high-resolution mid-infrared (MIR) dual-comb interferometer (DCI) using spectral interleaving. By generating electro-optic frequency combs (EOFCs) at 1550 nm via a dual-drive Mach-Zehnder modulator (DD-MZM) and employing injection locking to create a linearly swept lightwave spanning 18 GHz, we achieve gapless spectral interleaving. This swept lightwave serves as the seed for dual EOFCs with a 1 MHz repetition rate difference, which are subsequently converted to the MIR region (3.3 µm) via difference frequency generation (DFG). Experimental results demonstrate a dual-comb spectrum spanning 486 GHz with a 100 MHz resolution, validated by direct detection and demodulation.

In this paper, a non-resonant quartz-enhanced photoacoustic spectroscopy (NR-QEPAS) sensor is reported for the first time, to the best of our knowledge. The non-resonant photoacoustic cell (PAC) serves as the region where the photoacoustic effect occurs. NR-QEPAS offers several advantages, including flexible quartz tuning fork (QTF) positioning, frequency-matching-free operation, and simplified optical alignment. A self-designed T-head QTF was utilized as an acoustic wave transducer. The sound pressure characteristics of the non-resonant PAC were simulated using the finite element method. A near-infrared distributed feedback (DFB) diode laser with a wavelength of 1650.96 nm was selected as the excitation source. Methane (CH4) was chosen as the target gas to validate the designed sensor’s performance. The experimental results showed that the designed non-resonant PAC worked in the plane wave state, and the sound pressure in the cavity was nearly uniform. The minimum detection limit (MDL) of the designed NR-QEPAS sensor for CH4 detection could be 1.09 ppm (1 ppm = 10-6) when the average time was 760 s.

Mitochondrial dynamics critically regulate cellular aging. Two-photon nonlinear structured illumination microscopy (TP-SIM), a low-phototoxicity live-cell imaging technique, was employed to dynamically track mitochondrial changes in senescent H9C2 cardiomyocytes. System validation in COS7 cells achieved 82-nm resolution, threefold higher than conventional microscopy, and sustained 5-min dynamic imaging. Compared to normal cells, senescent cells exhibited fragmented mitochondria. TP-SIM further captured impaired mitochondrial fusion dynamics during senescence through continuous imaging, demonstrating its dual capability for subcellular-resolution visualization and prolonged organelle tracking in live cells.

In this paper, we propose a novel complex-valued hierarchical multi-fusion neural network (CHMFNet) for generating high-quality holograms. The proposed architecture builds upon a U-Net framework, incorporating a complex-valued multi-level perceptron (CMP) module that enhances complex feature representation through optimized convolutional operations and advanced activation functions, enabling effective extraction of intricate holographic patterns. The framework further integrates an innovative complex-valued hierarchical multi-fusion (CHMF) block, which implements multi-scale hierarchical processing and advanced feature fusion through its specialized design. This integration of complex-valued convolution and specialized CHMF design enables superior optical information representation, generating artifact-reduced high-fidelity holograms. The computational results demonstrate the superior performance of the proposed method, achieving an average peak signal-to-noise ratio (PSNR) of 34.11 dB and structural similarity index measure (SSIM) of 0.95, representing significant improvements over conventional approaches. Both numerical simulations and experimental validations confirm CHMFNet’s enhanced capability in hologram generation, particularly in terms of detail reproduction accuracy and overall image fidelity.

We experimentally demonstrate low-latency, high-capacity, and long-haul coherent transmission, recirculated through a self-fabricated, low-loss, and long-span 20 km nested antiresonant nodeless fiber (NANF) in the C-band. By leveraging wavelength division multiplexing (WDM), polarization multiplexing, probabilistic amplitude shaping technology, and low-complexity receiver-side digital signal processing (DSP), we achieve a record-breaking transmission capacity of 16.763 Tb/s over 1000 km for the first time, to the best of our knowledge. This achievement represents a significant step forward in high-capacity and long-haul optical communication based on NANF.

Dynamic polarization control remains a major challenge in fiber-integrated metasurfaces, with most research focusing on static polarization. Herein, we propose a dynamic polarization metasurface (DP-MS) integrated at the end of a large-mode fiber. The DP-MS decouples orthogonal polarization states, enabling precise rotation of focal spots in response to changes in the polarization direction of incident quasi-plane waves. Simulations show that the DP-MS can generate single, vortex, and multi-focal spots. By optimizing signal functions, multi-focal spots with precise control over number, position, distance, and synchronized rotation are achieved. This study has potential applications in optical manipulation and beam shaping.

This review begins by elucidating the rationale for selecting phase optical time domain reflectometry (Φ-OTDR) based on coherent detection as the subject of study, underscoring the necessity and significance of analyzing noise suppression methodologies. With the aid of in-phase/quadrature (I/Q) demodulation process for phase extraction, the review analyzes the impact of noise on phase-extracted results in coherent Φ-OTDR. Subsequently, nine specific implementable pathways to mitigate noise effects are explored within three categories: amplitude signal, phase signal, and the data itself. For each pathway, the review summarizes the advancements in noise suppression research, providing a typical implementation case to deepen the reader’s understanding of noise processing principles and techniques. Finally, the review not only identifies the shortcomings but also provides insights into noise processing in coherent Φ-OTDR, underlining the current limitations and suggesting potential avenues for future investigation.

We demonstrate a 200 m outdoor 2 × 2 multiple-input multiple-output (MIMO) terahertz (THz) communication system operating at 300 GHz with 200 Gb/s polarization-division multiplexed quadrature phase-shift keying (PDM-QPSK) transmission. We propose an iteratively pruned two-dimensional convolutional neural network (2D CNN) equalizer that adaptively captures polarization crosstalk and temporal nonlinearities through 2D convolution kernels. The system achieves a bit error rate (BER) below the hard-decision forward error correction (HD-FEC) threshold at a lower power of 6 dBm, while reducing the computational complexity by 30.2% compared to the iteratively pruned one-dimensional (1D) CNN approach. This enables high-capacity and energy-efficient operation in long-distance THz links.

We propose a method for reconstructing the distributed forward state of polarization (SOP) in single-mode fibers (SMFs) to solve the problem of an unpredictable blind box of SOP evolution in many applications such as fiber-optic parametric amplification systems. Using polarization-sensitive optical frequency domain reflectometry (POFDR) and a quaternion-based model to describe polarization changes, our approach achieves high spatial resolution and precision. By an improved iterative fitting algorithm, the mean square error (MSE) of forward SOP reconstruction for approximately 100 consecutive measurement points was reduced to below 0.1%. This method enables visualization of SOP dynamics along the fiber, offering critical insights for polarization-dependent systems.

With the rapid development of nanofabrication and computational technology, on-chip computational spectrometers enable miniaturized, high-resolution spectral analysis. However, visible light on-chip spectrometers still face significant challenges in performance and cost-effectiveness. This study presents an on-chip computational spectrometer using amorphous silicon (a-Si) metasurfaces. A strategy is employed that combines a genetic algorithm (GA) to assist in improving the spectral reconstruction algorithm, which effectively minimizes reconstruction errors and maximizes spectral resolution. The device achieves 1.5 nm resolution with 25 filter channels across a 300 nm bandwidth. Fabricated via complementary-metal-oxide-semiconductor (CMOS)-compatible processes, the spectrometer delivers high performance, compactness, and cost-effectiveness, showing great promise for miniaturized visible light spectral applications.

We have designed and fabricated a polarization light-emitting diode (LED) utilizing asymmetric nanograting metasurfaces, characterized by distinct grating structures on the top and bottom surfaces. Experimental results indicate a 34.66% improvement in polarization light extraction efficiency within ±60° relative to traditional metal-coated sapphire substrates. The measured average extinction ratio surpasses 21.62 dB within this angular range. By incorporating a half-wave plate function at the bottom through asymmetric nanograting metasurfaces, this LED design streamlines fabrication processes, reduces complexity, and enhances the efficiency of linearly polarized light. This innovative approach presents a promising solution for micro-LED failure analysis, advanced optical displays, communication systems, and photonic computing applications.

Microscale covert photonic barcodes demonstrate exceptional potential in anti-counterfeiting and information security applications due to their advanced security characteristics. However, the current methods suffer from spectral overlap and low concealment of security, restricting encoding capacity and requiring a high security level. These inherent drawbacks significantly restrict both the encoding capacity and the achievable security level. Here, we proposed a strategy to construct the high-security photonic barcodes via photomerization manipulation based on an excited-state intramolecular proton transfer (ESIPT) process in dye-doped whispering gallery mode (WGM) microcavities. The WGM microcavity is composed of highly polarized organic intramolecular charge-transfer (ICT) dye molecules, which have two cooperative gain states. Moreover, the light-manipulated covert photonic barcodes have further been obtained through an ESIPT energy-level process between the trans-excited state and cis-excited state. The WGM lasing spectrum constitutes the fingerprint of the corresponding microsphere, which can be modulated through tuning the dimensions of the microspheres. These results offer a promising route for exploring the light-manipulated security platform for advanced information anticounterfeiting.

A spectrum-tunable 650 nm semiconductor laser dense spectral beam combining (DSBC) system has been successfully realized for the first time, to the best of our knowledge. This system is constructed based on a dual-grating DSBC (DG-DSBC) module, which can realize DSBC with any preset spectrum width under ideal conditions. In this paper, three sets of spectrum-tunable examples are finally given. The combined spectra are stabilized at 4.89, 8.04, and 10.17 nm, with a maximum beam combining efficiency of about 88.27%. The brightness of this system is improved by more than 71% compared with that of the traditional DSBC structure.

We report continuous operation of stimulated Raman scattering at 1.9 µm wavelength based on hydrogen-filled anti-resonant hollow-core fibers (AR-HCFs) for the first time, to the best of our knowledge. Using a single-frequency fiber laser at 1 µm as the pump source, a Stokes laser with a maximum power over 25 W is measured in a 47 m nested type of AR-HCF filled with hydrogen gas at 10 bar pressure, corresponding to a power conversion efficiency of 40% and a quantum efficiency of 72.5%. Backward stimulated Raman scattering is observed at the same time and a maximum power of over 5 W is measured at a higher pressure of 30 bar. This work demonstrates the potential of gas-filled AR-HCF in high-power nonlinear wavelength conversion in the mid-infrared spectral region.

Spatiotemporal mode-locked (STML) fiber lasers have emerged as a novel platform for investigating spatiotemporal solitons and three-dimensional nonlinear phenomena. In this work, we report the generation of synchronous dual-wavelength STML noise-like square pulses in a few-mode fiber laser, characterized by distinct pulse durations at each wavelength. To further explore the experimental results, numerical simulations are conducted, where the mode-related and wavelength-related characteristics of the dual-wavelength noise-like pulses are revealed. It is found that different modes have distinct transient time-frequency characteristics, and a broader spectrum correlates with a longer duration of the pulse envelope and a shorter duration of the sub-pulses. These findings enhance our understanding of the underlying mechanisms and characteristics of noise-like pulses in STML fiber lasers for better exploration of their potential applications.

Airy wavepackets, distinguished by their unique self-accelerating, self-healing, and nondiffracting properties, have found extensive applications in particle manipulation, biomedical imaging, and material processing. Investigations into Airy waves have predominantly concentrated on either spatial or temporal dimensions, whereas studies on spatiotemporal Airy wavepackets have garnered less attention owing to the intricate nature of their generation systems. In this study, we present the generation of spatiotemporal Airy wavepackets by employing discrete frequency modulation and geometric phase modulation of pulses from a mode-locked fiber laser. The properties of Airy wavepackets are dictated by the imparted cubic frequency phase, geometric phase, and polarization state, resulting in controllable spatiotemporal profiles. The self-healing properties of spatiotemporal Airy wavepackets have been confirmed in both temporal and spatial dimensions, demonstrating substantial potential for applications in dynamic microscopy imaging and high-speed optical data transmission.

This work investigates spatial evolution characteristics during second-harmonic generation (SHG) through numerical and experimental study by employing a dual-pass Nd: YLF amplifier chain. Through simultaneous monitoring of conversion efficiency dynamics and beam profile evolution, we demonstrate that the spatial uniformity follows deterministic transformation patterns during nonlinear frequency conversion. Notably, optimization of beam uniformity was achieved at the fundamental power density of 0.478 GW/cm2 in our configuration, while maintaining conversion efficiency exceeding 85%.

Optical angular momentum (AM), comprising spin angular momentum (SAM) and orbital angular momentum (OAM), is crucial in various applications, yet its flexible control remains challenging. This study proposes, to our knowledge, a novel method for manipulating SAM and OAM using spherical wave illumination and the Λ-shaped spiral aperture. By adjusting the spherical wave’s convergence or divergence, the sign of SAM and OAM can be switched, while the geometric topological charge of the aperture transfers to the optical AM due to AM conservation. The method is theoretically analyzed, simulated, and experimentally validated, offering a compact platform applicable to photonic systems, particle manipulation, and encryption.

In this Letter, we propose an E-shaped hole metasurface leveraging bound states in the continuum (BICs) for perfect absorption and ultrasensitive refractive index sensing. We can achieve 98.9% optical absorption at 909 nm in the symmetric metasurface through a symmetry-protected BIC mode. It is found that there is a squared relationship between the Q-factor and the asymmetry factor. More importantly, we successfully activate the quasi-BIC (Q-BIC) mode in a symmetry-broken structure at an 8° light incidence angle. Both the symmetry-protected BIC and Q-BIC modes show 1000 nm/RIU (refractive index unit) sensitivity in air, while the latter outperforms in organic solutions with a figure of merit (FOM) of 463.9. This platform offers a versatile solution for ultra-narrowband photonics and high-precision biosensing applications.

We demonstrated that the epsilon-near-zero (ENZ) aluminum-doped zinc oxide (AZO) thin film exhibited ultrafast nonlinear optical response and efficient third-harmonic generation (THG) experimentally. The AZO film showed sub-picosecond response and broadband wavelength-dependent nonlinear absorption and refraction properties. In addition, the AZO thin film can produce efficient THG with an efficiency of 0.63 × 10-6 at the ENZ wavelength. The experimental results revealed the exceptional nonlinear optical behavior in the AZO thin film, and may provide insights for designing all-optical ultrafast optoelectronic devices.

This study demonstrates for the first time, to the best of our knowledge, that femtosecond laser-induced periodic surface structures (LIPSSs) enhance diamond’s visible-light transmittance. Using cylindrical-lens-shaped beams for high-speed scanning and secondary defocused low-energy laser treatment, uniform nanogratings with a period of 105 nm (12 µm × 24 µm) were fabricated within 3 s. Optimized scanning speeds and pulse energies improved structural quality, achieving up to 10% transmittance enhancement at 625–750 nm. This approach offers a novel strategy for anti-reflective diamond optoelectronic devices.

In this review, we address the emerging field of quantum photonic sensing leveraging the polarization degree of freedom. We briefly discuss the main aspects of treating polarization in quantum optics, and provide an overview of the main trends in the development of the field and the strategies to realize quantum-enhanced polarization-based sensing as well as a comprehensive survey of the main advancements in the field. We aim at promoting quantum approaches to the researchers in classical optical polarimetry as well as underscoring the sustainability and resourcefulness of the field for prospective applications and attracting the researchers in quantum optics to this new emerging field.

Single-photon detection (SPD) technologies have been applied to underwater optical imaging to overcome the strong attenuation of seawater. However, external photon noise, resulting from the natural light, hinders their further applications due to the extreme sensitivity of SPD and a weakly received optical signal. In this work, we performed noise-resistant underwater correlated biphoton imaging (CPI) to partly solve the influence of the external noise, through a home-built super-bunching laser generated by the stochastic nonlinear interaction between a picosecond laser and a photonic crystal fiber. Compared with a coherent laser, the probabilities of generated bundle N-photons (N ≥ 2) of the super-bunching laser have been enhanced by at least one order of magnitude, enabling CPI under weak light intensity. We experimentally demonstrated CPI with reasonable imaging contrast under the noise-to-signal ratio (NSR) up to 103, and the noise-resistant performance has been improved by at least two orders of magnitude compared to that of the single-photon imaging technology. We further achieved underwater CPI with good imaging contrast under NSR of 150, in a glass tank with a length of 10 m with Jerlov type III water (an attenuation coefficient of 0.176 m-1). These results break the limits of underwater imaging through classical coherent lasers and may offer many enhanced imaging applications through our super-bunching laser, such as long-range target tracking and deep-sea optical exploration under noisy environments.

Optical lattice clocks demonstrate advantages in metrology and frontier physics because of their high stability. Here, we present approaches to enhancing the stability by decoupling the noise related to the short-term and long-term stability. For the short-term stability, we optimize the clock laser by decoupling the frequency noise, and optimize each noise contribution individually until it is below the thermal noise limit. For the long-term stability, we introduce a method to decouple the instability caused by systematic effects. Having identified that the collision frequency shift was the main limiting factor in our systems, we thus optimized the atom number fluctuations in optical lattices. Through targeted optimization, we achieve a synchronous comparison of two clocks with an average stability of 3.2×10-16/τ and a long-term stability of 2.4 × 10-18 at 8000 s. This work provides an analytical framework for enhancing optical clock stability.