In this work, a dielectric metasurface composed of a silicon nanodisk etched with a square hole is proposed. By introducing C_4v symmetry breaking, the symmetry-protected bound states in the continuum (SP-BIC) is transformed into a quasi-BIC (Q-BIC), simultaneously inducing triple Fano resonances in the near-infrared light band corresponding to one dipole and two Q-BIC resonances. The characteristics of Q-BIC resonances are elucidated through multipole decomposition and near-field distribution analysis. Subsequently, monolayer graphene is integrated into the Si metasurface. The light field in the composite metasurface can be flexibly modulated by changing the Fermi level of graphene. This modulation enables optimal transmission with an enhancement of up to 252%, while the confined electromagnetic energy experiences a remarkable increase of about 1020%. Simulation results demonstrate that the Si–graphene composite metasurface exhibits a high refractive index sensitivity of 162 nm RIU⁻¹, accompanied by a figure of merit of 170.526 RIU⁻¹. This composite metasurface holds promise as a high-performance sensor in the near-infrared band and has potential for application in the fields of active tunable optical devices and biochemical sensing.

This study aimed to identify how different types of uniformity indicators function to evaluate the uniformity of lighting, such as the minimum-to-average ratio (Min:Avg), the coefficient of variance (CV), entropy uniformity (EU, introduced in the article), and a pattern vision-based indicator (U_HVS). Two experiments, a pair-wise comparison and a seven-point scale subjective evaluation, were performed to study the advantages and disadvantages of uniformity indicators. Thirty abstract patterns simulated from architectural illumination scenes were applied in the experiments. The results of the pair-wise comparison experiment reveal the following: the uniformity ranking and the sorting of uniformity indicator values obeyed a linear relationship; EU and U_HVS exhibited superior performance, with an adjusted R² of 0.96; the CV was satisfactory, with an adjusted R² of 0.91; and the Min:Avg was somewhat acceptable, with an adjusted R² of 0.70. The results of the seven-point scale subjective evaluation experiment demonstrated that ratings were linear with the Min:Avg, CV, and U_HVS, of which U_HVS was most consistent with human ratings of acceptability, with an adjusted R² of 0.96; the CV was acceptable, with an adjusted R² of 0.85; and the Min:Avg was the worst, with an adjusted R² of 0.71. The subjective evaluation rating and the EU indicator exhibited an exponential relationship, with an adjusted R² of 0.95, rather than a linear relationship. Different metrics have their own characteristics and may be suitable for different applications.

Several models on the circadian effect have been applied to indoor circadian lighting design, but applications in road lighting have not yet been clarified. Based on existing models and circadian research, we examined equivalent melanopic lux (EML), circadian light (CL_A), and circadian stimulus (CS) representing the circadian effect and the S/P ratio representing the mesopic vision effect, among a dataset of light sources at photopic adaptation illuminance values of 1, 3, 10, 30, and 100 lx. The results show that the S/P ratio correlates with EML and CS (or CL_A) much stronger than it correlates with color temperature. The EMLs of light sources are below 50 EML in mesopic vision, and the CSs of most light sources are below or around the threshold value of 0.05. We conclude that the circadian effect is not a significant issue in mesopic vision under most conditions and that optimization for mesopic efficiency is still a good strategy. There are quite a few light sources that may achieve both ideal mesopic efficiency and low CS. This work clarifies the circadian effect and mesopic vision effect performance of light sources in mesopic vision and will help guide choosing suitable light sources and optimization strategies for road lighting.

In this paper, ultrasensitive electrochemical determination of phosphate in water is achieved by using hydrophilic TiO ₂ modified glassy carbon electrodes for the first time. A detection limit of 0.1 μg L ⁻¹ is obtained, and a linear relationship (R ² = 0.99) between the phosphate concentration (ranging from 0.1 μg L ⁻¹ to 1 μg L ⁻¹ ) and the peak current was observed.

Spiking neural networks (SNNs) have emerged as a promising model for energy-efficient, event-driven processing of asynchronous event streams from Dynamic Vision Sensors (DVSs), a class of neuromorphic image sensors with microsecond-level latency and high dynamic range. Nevertheless, challenges persist in optimising training and effectively handling spatio-temporal complexity, which limits their potential for real-time applications on embedded sensing systems such as object tracking and recognition. Targeting this neuromorphic sensing pipeline, this paper proposes the Dynamic Tracking with Event Attention Spiking Network (DTEASN), a novel framework designed to address these challenges by employing a pure SNN architecture, bypassing conventional convolutional neural network (CNN) operations, and reducing GPU resource dependency, while tailoring the processing to DVS signal characteristics (asynchrony, sparsity, and polarity). The model incorporates two innovative, self-developed components: an event-driven multi-scale attention mechanism and a spatio-temporal event convolver, both of which significantly enhance spatio-temporal feature extraction from raw DVS events. An Event-Weighted Spiking Loss (EW-SLoss) is introduced to optimise the learning process by prioritising informative events and improving robustness to sensor noise. Additionally, a lightweight event tracking mechanism and a custom synaptic connection rule are proposed to further improve model efficiency for low-power, edge deployment. The efficacy of DTEASN is demonstrated through empirical results on event-based (DVS) object recognition and tracking benchmarks, where it outperforms conventional methods in accuracy, latency, event throughput (events/s) and spike rate (spikes/s), memory footprint, spike-efficiency (energy proxy), and overall computational efficiency under typical DVS settings. By virtue of its event-aligned, sparse computation, the framework is amenable to highly parallel neuromorphic hardware, supporting on- or near-sensor inference for embedded applications.