Search

Dynamic manipulation of terahertz (THz) beams plays an important role in THz application systems. The PB metasurface provides an effective scheme for space separation and deflection of the spin beam. However, mirror symmetry locking of the conjugated spin states severely limits the versatility of the device. In this work, we demonstrate a liquid crystal (LC) cascaded bilayer metasurface that includes an LC layer, anisotropic metasurface, and PB metasurface. By controlling anisotropy and polarization conversion effects, dynamic spin asymmetric transmission is realized. Meanwhile, two different dynamic energy distribution processes are realized between the L and R state with the corresponding deflection side. The results show that the device achieves a large angular spatial dispersion within the frequency-angle scanning range of ±35° to ±75° corresponding to the broadband range of 0.6–1.1 THz. Moreover, it achieves a spin beam spatial separation with a maximum proportion of energy distribution greater than 26 dB, and the active modulation rate in the energy distribution process reaches 98 %. This work provides a dynamic THz spin conversion and efficient large-angle beam scanning, with important potentials in wavelength/polarization division multiplexing and frequency-scanning antenna for large-capacity THz wireless communication, radar, and imaging systems.

Aiming at the coupling problem of target detection with few shot intensification, a few shot target detection algorithm based on decoupling is proposed focusing on high-value air targets as the research object. Firstly, a gradient adjustment layer is introduced into the regional candidate network to strengthen the regional candidate network and alleviate the task coupling problem. Secondly, the target detection head is disassembled into two branches, classification and regression, and a parameter free average attention module is added at the front end to alleviate the feature coupling problem. The proposed algorithm improves the detection performance of few shot target detection and enhances the detection ability of new classes. The experimental results show that the proposed algorithm performs best in the 1, 2, 3, 5 and 10 shot experiments, with average accuracy of 32. 5%, 35. 6%, 39. 6%, 41. 2% and 57. 4%, respectively. Compared with the two-stage fine- tuning method, the detection performance of the proposed algorithm is greatly improved, which solves the problem of the deliine of network detection ability under the coupling contradiction of few shot intensification, and improves the detection accuracy of few shot high-value air targets. [ABSTRACT FROM AUTHOR]

The task of camouflaged object segmentation (COS) is a challenging endeavor that entails the identification of objects that closely blend in with their surrounding background. Furthermore, the camouflaged object's obscure form and its subtle differentiation from the background present significant challenges during the feature extraction phase of the network. In order to extract more comprehensive information, thereby improving the accuracy of COS, we propose a diffusion model for a COS network that utilizes frequency domain information as auxiliary input, and we name it FreDiff. Firstly, we proposed a frequency auxiliary module (FAM) to extract frequency domain features. Then, we designed a Global Fusion Module (GFM) to make FreDiff pay attention to the global features. Finally, we proposed an Upsample Enhancement Module (UEM) to enhance the detailed information of the features and perform upsampling before inputting them into the diffusion model. Additionally, taking into account the specific characteristics of COS, we develop the specialized training strategy for FreDiff. We compared FreDiff with 17 COS models on the four challenging COS datasets. Experimental results showed that FreDiff outperforms or is consistent with other state-of-the-art methods under five evaluation metrics. [ABSTRACT FROM AUTHOR]

Camouflaged object detection (COD) is a challenging task that involves identifying objects that closely resemble their background. In order to detect camouflaged objects more accurately, we propose a diffusion model for the COD network called DMNet. DMNet formulates COD as a denoising diffusion process from noisy boxes to prediction boxes. During the training stage, random boxes diffuse from ground-truth boxes, and DMNet learns to reverse this process. In the sampling stage, DMNet progressively refines random boxes to prediction boxes. In addition, due to the camouflaged object's blurred appearance and the low contrast between it and the background, the feature extraction stage of the network is challenging. Firstly, we proposed a parallel fusion module (PFM) to enhance the information extracted from the backbone. Then, we designed a progressive feature pyramid network (PFPN) for feature fusion, in which the upsample adaptive spatial fusion module (UAF) balances the different feature information by assigning weights to different layers. Finally, a location refinement module (LRM) is constructed to make DMNet pay attention to the boundary details. We compared DMNet with other classical object-detection models on the COD10K dataset. Experimental results indicated that DMNet outperformed others, achieving optimal effects across six evaluation metrics and significantly enhancing detection accuracy. [ABSTRACT FROM AUTHOR]

The constraints of spatiotemporal heterogeneity and spatial resolution constitute two crucial challenges in the establishment of remote sensing inversion models. Spatiotemporal heterogeneity gives rise to an inadequate generalization capacity of remote sensing models, demanding extensive manual parameter adjustment for each model construction. This not only escalates the task's work intensity but also leads to unstable performance. The limited spatial resolution of remote sensing images leads to suboptimal inversion accuracy for sampling points influenced by mixed pixel effects. To tackle these problems, we take the case of non-photosensitive water quality parameter inversion in the narrow rivers of Longnan area. By integrating advanced Hyper-Parameter Optimization (HPO) techniques, such as Optuna from machine learning, an inversion model was developed, incorporating the bands of Sentinel-2 and Sentinel-3 as model features. Among these features, bands with lower spatial resolution are employed to furnish surrounding information, thereby enhancing the inversion accuracy. The research outcomes demonstrate that: 1) The model constructed based on the HPO method, Optuna, attained favourable inversion results, with R2 values of 0.68, 0.77, 0.35, and 0.60 for Total Nitrogen (TN), Total Phosphorus (TP), Ammonia Nitrogen (NH3-N), and Chemical Oxygen Demand (COD), respectively. 2) The fusion of Sentinel-2 and Sentinel-3 data enhanced the inversion accuracy compared to using them separately, highlighting the considerable significance of multi-source data fusion methods in improving inversion accuracy. This research fills a void in the remote sensing inversion domain and lays the groundwork for future endeavours. [ABSTRACT FROM AUTHOR]

Deep learning models are highly susceptible to adversarial samples, and even minuscule image perturbations that are not perceptible to the naked eye can disable well-trained deep learning models. Recent research indicates that these perturbations can exist in the physical world. This paper provides insight into physical adversarial attacks on deep learning object detection models, clarifying the concept of physical adversarial attack and outlining the general process of such attacks on object detection. According to the different attack tasks, a series of physical adversarial attack methods against object detection networks in recent years are reviewed from vehicle detection and pedestrian detection. Other attacks against target detection models, other attack tasks and other attack methods are briefly introduced. The current challenges of physical adversarial attack are discussed, the limitations of adversarial training are leaded out, and future development directions and application prospect are suggested. [ABSTRACT FROM AUTHOR]

This study investigated the transcriptomic mechanisms underlying melatonin accumulation and the enhancement of salt tolerance in hull-less barley seeds subjected to zinc sulphate stress. Following zinc sulphate treatment, hull-less barley seeds demonstrated increased melatonin accumulation and improved salt tolerance. Through transcriptome analysis, the study compared gene expression alterations in seeds (using the first letter of seed, this group is marked as 'S'), seeds treated with pure water (as the control group, is marked as 'C'), and germinated seeds exposed to varying concentrations of zinc sulphate (0.2 mM and 0.8 mM, the first letter of zinc sulphate, 'Z', is used to mark groups 'Z1' and 'Z2'). The analysis revealed that 8176, 759, and 622 differentially expressed genes (DEGs) were identified in the three comparison groups S.vs.C, C.vs.Z1, and C.vs.Z2, respectively. Most of the DEGs were closely associated with biological processes, including oxidative-stress response, secondary metabolite biosynthesis, and plant hormone signaling. Notably, zinc sulphate stress influenced the expression levels of Tryptophan decarboxylase 1 (TDC1), Acetylserotonin O-methyltransferase 1 (ASMT1), and Serotonin N-acetyltransferase 2 (SNAT2), which are key genes involved in melatonin synthesis. Furthermore, the expression changes of genes such as Probable WRKY transcription factor 75 (WRKY75) and Ethylene-responsive transcription factor ERF13 (EFR13) exhibited a strong correlation with fluctuations in melatonin content. These findings contribute to our understanding of the mechanisms underlying melatonin enrichment in response to zinc sulphate stress. [ABSTRACT FROM AUTHOR]

Geospatial data, especially remote sensing (RS) data, are of significant importance for public services and production activities. Expertise is critical in processing raw data, generating geospatial information, and acquiring domain knowledge and other remote sensing applications. However, existing geospatial service platforms are more oriented towards the professional users in the implementation process and final application. Building appropriate geographic applications for non-professionals remains a challenge. In this study, a geospatial data service architecture is designed that links desktop geographic information system (GIS) software and cloud-based platforms to construct an efficient user collaboration platform. Based on the scalability of the platform, four web apps with different themes are developed. Data in the fields of ecology, oceanography, and geology are uploaded to the platform by the users. In this pilot phase, the gap between non-specialized users and experts is successfully bridged, demonstrating the platform's powerful interactivity and visualization. The paper finally evaluates the capability of building spatial data infrastructures (SDI) based on GeoNode and discusses the current limitations. The support for three-dimensional data, the improvement of metadata creation and management, and the fostering of an open geo-community are the next steps. [ABSTRACT FROM AUTHOR]

AI drawing tools, which rely on high-quality prompts to guide and refine the creative process, have the potential to enhance efficiency and quality during the conceptual design stage. However, addressing the inherent accuracy issues of AI is crucial. This study explores the efficiency and usability of integrating AI drawing with prompt optimisation tools and introduces a GPT-based prompt optimisation tool for validation. Experiment results indicate that, during the divergent thinking stage of concept design, both junior and senior designers using this tool reduced their average time consumption by 35.6% and 21.6%, respectively. Expert evaluations of the generated images reveal higher selection rates for images produced after prompt optimisation with this tool compared to those generated using natural language. These findings underscore the tool's potential to boost productivity and improve the usability of generated images, thereby expanding designers’ imaginative boundaries and enhancing their ability to leverage AI for streamlined workflows. [ABSTRACT FROM AUTHOR]

Glucose‐derived carbon‐decorated magnetic microspheres were synthesized by an easy hydrothermal carbonization method and used as a high‐efficiency adsorbent to extract bisphenols in water and tea drinks. The as‐prepared carbon‐decorated magnetic microspheres had a well‐defined core–shell structure with a shell thickness of about 5 nm. The microspheres possessed high saturation magnetization at 60.8 emu/g and excellent chemical stability in aqueous solution. The experimental parameters affecting the extraction efficiency, including extraction time, pH, adsorbent dosage, desorption solvents, desorption time, and solution volume were evaluated. Electrostatic and π–π interactions were the major driving forces during extraction. Overall, a new magnetic solid‐phase extraction method of determining bisphenols was developed on the basis of as‐prepared magnetic microspheres. The method had a wide linear range, low limits of detection (0.03–0.10 µg/L), and high recoveries (85.4–104.6%). [ABSTRACT FROM AUTHOR]