In this paper, multiple Fano resonances preferred in the refractive index sensing area are achieved based on sub-wavelength metal-insulator-metal (MIM) waveguides. Two slot cavities, which are placed between or above the MIM waveguides, can support the bright modes or the dark modes, respectively. Owing to the mode interferences, dual Fano resonances with obvious asymmetrical spectral responses are achieved. High sensitivity and high figure of merit are investigated by using the finite-difference time-domain (FDTD) method. In view of the development of chip-scale integrated photonics, two extra slot cavities are successively added to the structure, and consequently, three and four ultra-sharp Fano peaks with considerable performances are obtained, respectively. It is believed that this proposed structure can find important applications in the on-chip optical sensing and optical communication areas.

The micro-structure with submicron-scale precision so far as to minimum feature size of 1.5 μm can be achieved at faster fabrication speed. This rapid prototyping fabrication method take advantage of combining of maskless lithography technique based on digital mirror device (DMD) and photo-curable resin which has a broad range of use in three-dimensional printing (3D Printing). To illustrate the feasibility of this method, a kind of Dammann grating was chosen to be fabricated. By a series of experiments and optimization designs, we analyzed the characteristics of the photo-curable resin, quantified the dose modulation of lithography and redesigned the mask pattern. Consequently, the Dammann grating with 50% diffraction efficiency was fabricated successfully which can not only guarantee the precision and but also keep up the fabrication speed. This work demonstrates the potential of this method to rapidly and directly manufacturing binary-optics elements or structure in nano-scale based on photo-curable resin and DMD-based maskless lithography.

Object detection has become a crucial technology in intelligent vision systems, enabling automatic detection of target objects. While most detectors perform well on open datasets, they often struggle with small‐scale objects. This is due to the traditional top‐down feature fusion methods that weaken the semantic and location information of small objects, leading to poor classification performance. To address this issue, we propose a novel feature pyramid network, the adaptive learnable feature pyramid network (ALFPN). Our approach features an adaptive feature inspection that incorporates learnable fusion coefficients in the fusion of different levels of feature layers, aiding the network in learning features with less noise. In addition, we construct a context‐aligned supervisor that adjusts the feature maps fused at different levels to avoid scaling‐related offset effects. Our experiments demonstrate that our method achieves state‐of‐the‐art results and is highly robust for the small object detection on the TT‐100K, PASCAL VOC, and COCO datasets. These findings indicate that a model’s ability to extract discriminant features is positively correlated with its performance in detecting small objects.

A two-port beam splitter is presented to obtain the 50/50 output based on the two-layer grating. The previously reported beam splitter grating can work only at the special prescribed incident angle. The deviation of incident angle introduces the effect on the splitting ratio uniformity. The proposed two-layer beam splitter grating is aimed to obtain the broad incident angular bandwidth. To optimize the beam splitter, a simplified modal method and rigorous coupled-wave analysis are employed to analyze the propagation process and to calculate the efficiency. Based on the optimized grating parameters, the 50/50 output can be obtained with the polarization-independent property. Most importantly, good splitting ratio uniformity can be exhibited within the broad angular bandwidth for both TE and TM polarizations. The polarization-independent property and the broad incident angular bandwidth are significant for practical applications.

Efficient dense reconstruction of objects or scenes has substantial practical implications, which can be applied to different 3D tasks (for example, robotics and autonomous driving). However, because of the expensive hardware required and the overall complexity of the all-around scenarios, efficient dense reconstruction using lightweight multi-view stereo methods has received much attention from researchers. The technological challenge of efficient dense reconstruction is maintaining low memory usage while rapidly and reliably acquiring depth maps. Most of the current efficient multi-view stereo (MVS) methods perform poorly in efficient dense reconstruction, this poor performance is mainly due to weak generalization performance and unrefined object edges in the depth maps. To this end, we propose EMO-MVS, which aims to accomplish multi-view stereo tasks with high efficiency, which means low-memory consumption, high accuracy, and excellent generalization performance. In detail, we first propose an iterative variable optimizer to accurately estimate depth changes. Then, we design a multi-level absorption unit that expands the receptive field, which efficiently generates an initial depth map. In addition, we propose an error-aware enhancement module, enhancing the initial depth map by optimizing the projection error between multiple views. We have conducted extensive experiments on challenging datasets Tanks and Temples and DTU, and also performed a complete visualization comparison on the BlenedMVS validation set (which contains many aerial scene images), achieving promising performance on all datasets. Among the lightweight MVS methods with low-memory consumption and fast inference speed, our F-score on the online Tanks and Temples intermediate benchmark is the highest, which shows that we have the best competitiveness in terms of balancing the performance and computational cost.