Search Results (6)

Hyperspectral sparse unmixing (SU) networks have recently received considerable attention due to their model hyperspectral images (HSIs) with a priori spectral libraries and to capture nonlinear features through deep networks. This method effectively avoids errors associated with endmember extraction, and enhances the unmixing performance via nonlinear modeling. However, two major challenges remain: the use of large spectral libraries with high coherence leads to computational redundancy and performance degradation; moreover, certain feature extraction models, such as Transformer, while exhibiting strong representational capabilities, suffer from high computational complexity. To address these limitations, this paper proposes a hyperspectral unmixing dual-branch network integrating an adaptive hop-aware GCN and neighborhood offset Mamba that is termed DGMNet. Specifically, DGMNet consists of two parallel branches. The first branch employs the adaptive hop-neighborhood-aware GCN (AHNAGC) module to model global spatial features. The second branch utilizes the neighborhood spatial offset Mamba (NSOM) module to capture fine-grained local spatial structures. Subsequently, the designed Mamba-enhanced dual-stream feature fusion (MEDFF) module fuses the global and local spatial features extracted from the two branches and performs spectral feature learning through a spectral attention mechanism. Moreover, DGMNet innovatively incorporates a spectral-library-pruning mechanism into the SU network and designs a new pruning strategy that accounts for the contribution of small-target endmembers, thereby enabling the dynamic selection of valid endmembers and reducing the computational redundancy. Finally, an improved ESS-Loss is proposed, which combines an enhanced total variation (ETV) with an $l_{1/2}$ sparsity constraint to effectively refine the model performance. The experimental results on two synthetic and five real datasets demonstrate the effectiveness and superiority of the proposed method compared with the state-of-the-art methods. Notably, experiments on the Shahu dataset from the Gaofen-5 satellite further demonstrated DGMNet’s robustness and generalization. Full article

The mass of redundant and irrelevant data in network traffic brings serious challenges to intrusion detection, and feature selection can effectively remove meaningless information from the data. Most current filtered and embedded feature selection methods use a fixed threshold or ratio to determine the number of features in a subset, which requires a priori knowledge. In contrast, wrapped feature selection methods are computationally complex and time-consuming; meanwhile, individual feature selection methods have a bias in evaluating features. This work designs an ensemble-based automatic feature selection method called EAFS. Firstly, we calculate the feature importance or ranks based on individual methods, then add features to subsets sequentially by importance and evaluate subset performance comprehensively by designing an NSOM to obtain the subset with the largest NSOM value. When searching for a subset, the subset with higher accuracy is retained to lower the computational complexity by calculating the accuracy when the full set of features is used. Finally, the obtained subsets are ensembled, and by comparing the experimental results on three large-scale public datasets, the method described in this study can help in the classification, and also compared with other methods, we discover that our method outperforms other recent methods in terms of performance. Full article

Recent advances in Surface Plasmon Resonance (SPR) technologies have shown the possibility of transmission enhancement of localized modes propagating through sub-diffraction wide slits and apertures, resulting in the strong near-field focusing of metallic planar nanostructures. This work presents a new approach to the fabrication of high-resolution near-field optical probes using 3D lithography in combination with numerical finite difference time domain (FDTD) simulations. A narrow 500 nm depth of field focus area was observed both by numerical analysis and near field scanning optical microscopy (NSOM) measurements. Further research and optimization are planned in order to achieve subwavelength focal regions and increased signal intensities. Full article

In winemaking, clarification and stabilization are the processes by which insoluble matter suspended in the wine (called lees) is removed before bottling. The light lees represent 2–4% of the total wine volume. Under certain circumstances, resuspension of lees may occur. The resuspension of lees has been attributed to temperature variations between the wine stored in tanks and the environment of the cellar. From in situ, laboratory-scale studies involving laser tomography techniques, it was shown that low (positive or negative) thermal gradients between a wine tank containing light lees and its external environment induce mass transfer by natural convection. To extrapolate these findings to full-scale tanks, an Eulerian-Eulerian multiphase CFD model was applied to simulate the two-phase flow behavior as a function of temperature variations on a 24–h cycle. Numerical temperature and time-dependent flow patterns of both wine and lees confirm that low thermal gradients induce sufficient fluid energy to resuspend the lees, thus showing that the laboratory results can be extrapolated to full-scale tanks. Full article

A feasibility analysis is performed for the development and integration of a near-field scanning optical microscope (NSOM) tip–photodetector operating in the visible wavelength domain of an atomic force microscope (AFM) cantilever, involving simulation, processing, and measurement. The new tip–photodetector consists of a platinum–silicon truncated conical photodetector sharing a subwavelength aperture, and processing uses advanced nanotechnology tools on a commercial silicon cantilever. Such a combined device enables a dual-mode usage of both AFM and NSOM measurements when collecting the reflected light directly from the scanned surface, while having a more efficient light collection process. In addition to its quite simple fabrication process, it is demonstrated that the AFM tip on which the photodetector is processed remains operational (i.e., the AFM imaging capability is not altered by the process). The AFM–NSOM capability of the processed tip is presented, and preliminary results show that AFM capability is not significantly affected and there is an improvement in surface characterization in the scanning proof of concept. Full article

The size of the V-defects in the GaN/InGaN-based quantum wells blue light-emitting diode (LED) was intentionally modified from 50 nm to 300 nm. High resolution photoluminescence and electroluminescence of a single large V-defect were investigated by near-field scanning optical microscopy. The current distribution along the {10-11} facets of the large defect was measured by conductive atomic force microscopy. Nearly 20 times the current injection and dominant emission from bottom quantum wells were found in the V-defect compared to its vicinity. Such enhanced current injection into the bottom part of quantum wells through V-defect results in higher light output power. Reduced external quantum efficiency droops were achieved due to more uniform carrier distribution. The un-encapsulated fabricated chip shows light output power of 172.5 mW and 201.7 mW at 400 mA, and external quantum efficiency drop of 22.3% and 15.4% for the sample without and with large V-defects, respectively. Modified V-defects provide a simple and effective approach to suppress the efficiency droop problem that occurs at high current injection, while improving overall quantum efficiency. Full article