Spectral Features Research Articles

Progressive High-Frequency Reconstruction for Pan-Sharpening with Implicit Neural Representation

Pan-sharpening aims to leverage the high-frequency signal of the panchromatic (PAN) image to enhance the resolution of its corresponding multi-spectral (MS) image. However, deep neural networks (DNNs) tend to prioritize learning the low-frequency components during the training process, which limits the restoration of high-frequency edge details in MS images. To overcome this limitation, we treat pan-sharpening as a coarse-to-fine high-frequency restoration problem and propose a novel method for achieving high-quality restoration of edge information in MS images. Specifically, to effectively obtain fine-grained multi-scale contextual features, we design a Band-limited Multi-scale High-frequency Generator (BMHG) that generates high-frequency signals from the PAN image within different bandwidths. During training, higher-frequency signals are progressively injected into the MS image, and corresponding residual blocks are introduced into the network simultaneously. This design enables gradients to flow from later to earlier blocks smoothly, encouraging intermediate blocks to concentrate on missing details. Furthermore, to address the issue of pixel position misalignment arising from multi-scale features fusion, we propose a Spatial-spectral Implicit Image Function (SIIF) that employs implicit neural representation to effectively represent and fuse spatial and spectral features in the continuous domain. Extensive experiments on different datasets demonstrate that our method outperforms existing approaches in terms of quantitative and visual measurements for high-frequency detail recovery.

Hyperspectral Image Reconstruction via Combinatorial Embedding of Cross-Channel Spatio-Spectral Clues

Existing learning-based hyperspectral reconstruction methods show limitations in fully exploiting the information among the hyperspectral bands. As such, we propose to investigate the chromatic inter-dependencies in their respective hyperspectral embedding space. These embedded features can be fully exploited by querying the inter-channel correlations in a combinatorial manner, with the unique and complementary information efficiently fused into the final prediction. We found such independent modeling and combinatorial excavation mechanisms are extremely beneficial to uncover marginal spectral features, especially in the long wavelength bands. In addition, we have proposed a spatio-spectral attention block and a spectrum-fusion attention module, which greatly facilitates the excavation and fusion of information at both semantically long-range levels and fine-grained pixel levels across all dimensions. Extensive quantitative and qualitative experiments show that our method (dubbed CESST) achieves SOTA performance. Code for this project is at: https://github.com/AlexYangxx/CESST.

A reanalysis of the LHS 1140 b atmosphere observed with the Hubble Space Telescope

ABSTRACT The super-Earth LHS 1140 b is an interesting target for atmospheric observations since it is close to the habitable zone of its star and falls in the gap of the radius distribution of small exoplanets, in the region thought to correspond to the transition between planets with and without atmospheres. Observations of the primary transit with WFC3 onboard of the Hubble Space Telescope (HST) revealed a modulation in the planet transmission spectrum compatible with the presence of water; however, this modulation may be also due to stellar activity-related phenomena. Here, we present a detailed analysis of the WFC3/HST observations to probe the nature of this modulation and to understand if it can be attributable to the presence of unocculted spots on the stellar surface. Our analysis strongly suggests that LHS1140 is a rather quiet star with subsolar [Fe/H] and enriched in α elements. Therefore, we rule out the possibility that the planetary spectrum is affected by the presence of spots and faculae. This analysis shows the importance of a proper modelling of the stellar spectrum when analysing transit observations. Finally, we modelled the planetary atmosphere of LHS1140 b to retrieve its atmospheric composition. However, the low resolution and the narrow spectral range of HST observations prevented us from definitively determining whether the spectral features are attributable to the presence of water or of other molecules in the planetary atmosphere.

Machine learning assisted Raman spectroscopy: A viable approach for the detection of microplastics

The accumulation of microplastics (MPs) resulting from disposal of plastic waste into water sources, poses a significant threat to aquatic organisms. These are readily ingested by organisms, leading to the accumulation of harmful substances, disrupting their biological processes. Current methods for identifying microplastics have notable drawbacks, including low resolution, extended imaging time, and restricted particle size analysis. Integrating Raman spectroscopy with machine learning (ML) proves to be an effective approach for identifying and classifying MPs, especially in scenarios where they are found in environmental media or mixed with various types. Machine learning (ML) can be vital tool in assisting Raman analysis, owing to its robust feature extraction capabilities. This comprehensive review outlined the utilization of various machine learning techniques in conjunction with Raman spectral features for diverse investigations related to microplastics. The methodologies discussed encompass Principal Component Analysis, K-Nearest Neighbour, Random Forest, Support Vector Machine, and various deep learning algorithms.

Spectral boundary detecting model: A promising tool for adaptive mode extraction and machinery fault diagnosis

Extraction of weak transients is vital for realizing the early machinery fault diagnosis. However, a significant challenge lies in an efficient determination of the spectral boundaries of expected modes for excluding interferential information. Through a new perspective exploration on the fundamental principle of variational mode decomposition, we find a relevance between the filtering structure of variational model and the shape of fault-induced features in spectral domain. Then, we propose a spectral boundary detecting model (SBDM) to identify fault-induced transients efficiently. SBDM comprises three steps. First, spectral structural features of analysis signal are quickly detected based on the found relevance. Second, spectral boundaries with expected modes are calculated based on the located spectral structural features. Third, a filter bank designed on the spectral boundaries is used to extract the expected modes. Furthermore, our SBDM is extended by two observed valuable phenomena to obtain a comprehensive result for downstream diagnosis tasks. Finally, an extended SBDM-based diagnosis procedure is presented for the machinery fault identification. Two experimental cases have validated that our SBDM outperforms eight advanced methods in the machinery fault diagnosis.

Vegetation Classification and Evaluation of Yancheng Coastal Wetlands Based on Random Forest Algorithm from Sentinel-2 Images

The identification of wetland vegetation is essential for environmental protection and management as well as for monitoring wetlands’ health and assessing ecosystem services. However, some limitations on vegetation classification may be related to remote sensing technology, confusion between plant species, and challenges related to inadequate data accuracy. In this paper, vegetation classification in the Yancheng Coastal Wetlands is studied and evaluated from Sentinel-2 images based on a random forest algorithm. Based on consistent time series from remote sensing observations, the characteristic patterns of the Yancheng Coastal Wetlands were better captured. Firstly, the spectral features, vegetation indices, and phenological characteristics were extracted from remote sensing images, and classification products were obtained by constructing a dense time series using a dataset based on Sentinel-2 images in Google Earth Engine (GEE). Then, remote sensing classification products based on the random forest machine learning algorithm were obtained, with an overall accuracy of 95.64% and kappa coefficient of 0.94. Four indicators (POP, SOS, NDVIre, and B12) were the main contributors to the importance of the weight analysis for all features. Comparative experiments were conducted with different classification features. The results show that the method proposed in this paper has better classification.

Forward-predictive SERS-based chemical taxonomy for untargeted structural elucidation of epimeric cerebrosides

Achieving untargeted chemical identification, isomeric differentiation, and quantification is critical to most scientific and technological problems but remains challenging. Here, we demonstrate an integrated SERS-based chemical taxonomy machine learning framework for untargeted structural elucidation of 11 epimeric cerebrosides, attaining >90% accuracy and robust single epimer and multiplex quantification with <10% errors. First, we utilize 4-mercaptophenylboronic acid to selectively capture the epimers at molecular sites of isomerism to form epimer-specific SERS fingerprints. Corroborating with in-silico experiments, we establish five spectral features, each corresponding to a structural characteristic: (1) presence/absence of epimers, (2) monosaccharide/cerebroside, (3) saturated/unsaturated cerebroside, (4) glucosyl/galactosyl, and (5) GlcCer or GalCer’s carbon chain lengths. Leveraging these insights, we create a fully generalizable framework to identify and quantify cerebrosides at concentrations between 10−4 to 10−10 M and achieve multiplex quantification of binary mixtures containing biomarkers GlcCer24:1, and GalCer24:1 using their untrained spectra in the models.

Using the automated HYPERNETS hyperspectral system for multi-mission satellite ocean colour validation in the Río de la Plata, accounting for different spatial resolutions

Validation of water reflectance using in situ data is essential to ensure the quality of ocean colour satellite-derived products useful for water quality monitoring, like turbidity and chlorophyll-a concentration. Since December 2021, the HYPERNETS automated hyperspectral system has been collecting data in the optically complex and highly turbid waters of the Río de la Plata, an ideal scenario for testing atmospheric correction algorithms’ performance. The site, located 60 km south of Buenos Aires (Argentina), is described in relation to the water reflectance spectral features and variability using high spatial resolution imagery and a methodology is proposed to objectively select a sensor-specific location of a reference pixel for satellite validation. Six months of data is used to evaluate surface water reflectance operational products from multi-spectral systems like Landsat 8&amp;amp;9/OLI (L89/OLI), Sentinel-2/MSI (S2/MSI) &amp;amp; Sentinel-3/OLCI (S3/OLCI), and PlanetScope SuperDoves (PS/SD), and also non standard products for Aqua/MODIS (Aqua/MODIS) and SNPP&amp;amp;JPSS1/VIIRS (SJ/VIIRS) missions. Moreover, the standard surface water reflectance product from the hyperspectral PRISMA mission could also be evaluated. The matchups show general good results when in situ measurements are compared to L2 standard products of high spatial resolution sensors that use land-based atmospheric correction approach, if sun glint contamination is avoided. Low mean relative percentage difference was found for S2/MSI (2.45%) and L89/OLI (−3.52%), but higher for PS/SD (30.7%). In turn, S3/OLCI medium resolution also showed low mean relative differences (2.31%), while SJ/VIIRS and Aqua/MODIS showed larger and negative differences (−16.35 for SJ/VIIRS and −35.6% for Aqua/MODIS) which showed a clear increase towards the shortest blue bands. The results show the great potential of the HYPERNETS automated system to provide high quality and quantity of data for validation of satellite data at all visible and near infrared (VNIR, 400–900 nm) wavelengths in a multi-mission perspective.

Spectra of phosphorus ions for astrophysical modeling: P I–P XV

Phosphorus (P), a basic element of life, has been a least studied element due to its poor presence in astrophysical spectra. However, search for the P lines has increased considerably with discoveries of exoplanets and are being detected by high resolution and sophisticated astronomical observatories, e.g., James Webb Space Telescope (JWST). JWST may provide a clue for life with detection of P in its infrared (IR) region. Identification of the element and analysis of the observed spectra will require high accuracy data for atomic processes that produces lines and their predicted features. The present study focuses on these needs and reports systematically regions of wavelengths, from X-ray to IR, that show prominent lines by the 15 individual ionization stages of phosphorus, P I–P XV for the first time. We present large amount of relevant atomic data for energies, transition parameters, and lifetimes obtained in relativistic Breit–Pauli approximation using the R-matrix method and atomic structure program SUPERSTRUCTURE. Our spectral features for the 15 ions, P I–P XV, predict strengths of lines in various wavelength regions. They show dominance of P I and P II in the IR region and other ions in the ultraviolet and optical regions often stretching to IR in the continuum. For determination of accuracy, we have made extensive comparisons of our atomic data with available experimental and theoretical values. Based on these, our results and features are expected to provide precise plasma diagnostics and astrophysical modeling.

Advancing Hyperspectral Image Analysis with CTNet: An Approach with the Fusion of Spatial and Spectral Features.

Hyperspectral image classification remains challenging despite its potential due to the high dimensionality of the data and its limited spatial resolution. To address the limited data samples and less spatial resolution issues, this research paper presents a two-scale module-based CTNet (convolutional transformer network) for the enhancement of spatial and spectral features. In the first module, a virtual RGB image is created from the HSI dataset to improve the spatial features using a pre-trained ResNeXt model trained on natural images, whereas in the second module, PCA (principal component analysis) is applied to reduce the dimensions of the HSI data. After that, spectral features are improved using an EAVT (enhanced attention-based vision transformer). The EAVT contained a multiscale enhanced attention mechanism to capture the long-range correlation of the spectral features. Furthermore, a joint module with the fusion of spatial and spectral features is designed to generate an enhanced feature vector. Through comprehensive experiments, we demonstrate the performance and superiority of the proposed approach over state-of-the-art methods. We obtained AA (average accuracy) values of 97.87%, 97.46%, 98.25%, and 84.46% on the PU, PUC, SV, and Houston13 datasets, respectively.

Therapeutic Implications of Voiced Sounds in Children with Communication Needs

Children with hearing impairments, who do not receive auditory feedback eventually develop speech impairments. Children with hearing impairments lack the ability to talk even by having normal speech production mechanisms. The hearing-impaired child's articulation, accuracy, stress, and intonation patterns are all impacted even if he/she attempts to speak by visualizing lip movements because they are unable to distinguish vowels and consonants with tongue movement concealed in the mouth, nor they are able to understand variations in pitch or intensity of speech. Individuals with hearing impairments may receive auditory, tactile, or visual input, contingent on the degree of their disability. This research is dedicated for developing preliminary systems for teaching. These systems can be created using feedback of articulatory parameters or feedback of acoustic parameters, such as fundamental frequency, speech intensity, and spectral features. Data from speech language pathologists were collected from 20 children with and without speech impairments, aged 5 to 8 using picture naming tasks. The statistical analysis on voiced samples /pha/, /mha/, /tha/ relived the significant variation between children with and without hearing impairments.

Constant-sized self-tests for maximally entangled states and single projective measurements

Self-testing is a powerful certification of quantum systems relying on measured, classical statistics. This paper considers self-testing in bipartite Bell scenarios with small number of inputs and outputs, but with quantum states and measurements of arbitrarily large dimension. The contributions are twofold. Firstly, it is shown that every maximally entangled state can be self-tested with four binary measurements per party. This result extends the earlier work of Mančinska-Prakash-Schafhauser (2021), which applies to maximally entangled states of odd dimensions only. Secondly, it is shown that every single binary projective measurement can be self-tested with five binary measurements per party. A similar statement holds for self-testing of projective measurements with more than two outputs. These results are enabled by the representation theory of quadruples of projections that add to a scalar multiple of the identity. Structure of irreducible representations, analysis of their spectral features and post-hoc self-testing are the primary methods for constructing the new self-tests with small number of inputs and outputs.

Forward modelling low-spectral-resolution Cassini/CIRS observations of Titan

The Composite InfraRed Spectrometer (CIRS) instrument onboard the Cassini spacecraft performed 8.4 million spectral observations of Titan at resolutions between 0.5–15.5 cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document}. More than 3 million of these were acquired at a low spectral resolution (SR) (13.5–15.5 cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document}), which have excellent spatial and temporal coverage in addition to the highest spatial resolution and lowest noise per spectrum of any of the CIRS observations. Despite this, the CIRS low-SR dataset is currently underused for atmospheric composition analysis, as spectral features are often blended and subtle compared to those in higher SR observations. The vast size of the dataset also poses a challenge as an efficient forward model is required to fully exploit these observations. Here, we show that the CIRS FP3/4 nadir low-SR observations of Titan can be accurately forward modelled using a computationally efficient correlated-k\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\varvec{k}$$\\end{document} method. We quantify wavenumber-dependent forward modelling errors, with mean 0.723 nW cm-2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-2}}\\,$$\\end{document}sr-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document}/cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document} (FP3: 600–890 cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document}) and 0.248 nW cm-2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-2}}\\,$$\\end{document}sr-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}\\,$$\\end{document}/ cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document} (FP4: 1240–1360 cm-1\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{\\varvec{-1}}$$\\end{document}), that can be used to improve the rigour of future retrievals. Alternatively, in cases where more accuracy is required, we show observations can be forward modelled using an optimised line-by-line method, significantly reducing computation time.

Near-infrared spectroscopy based on colorimetric sensor array coupled with convolutional neural network detecting zearalenone in wheat

Wheat is a vital global cereal crop, but its susceptibility to contamination by mycotoxins can render it unusable. This study explored the integration of two novel non-destructive detection methodologies with convolutional neural network (CNN) for the identification of zearalenone (ZEN) contamination in wheat. Firstly, the colorimetric sensor array composed of six selected porphyrin-based materials was used to capture the olfactory signatures of wheat samples. Subsequently, the colorimetric sensor array, after undergoing a reaction, was characterized by its near-infrared spectral features. Then, the CNN quantitative analysis model was proposed based on the data, alongside the establishment of traditional machine learning models, partial least squares regression (PLSR) and support vector machine regression (SVR), for comparative purposes. The outcomes demonstrated that the CNN model had superior predictive performance, with a root mean square error of prediction (RMSEP) of 40.92 μ g ∙ kg−1 and a coefficient of determination on the prediction (RP2) of 0.91. These results affirmed the potential of integrating colorimetric sensor array with near-infrared spectroscopy in evaluating the safety of wheat and potentially other grains. Moreover, CNN can have the capacity to autonomously learn and distill features from spectral data, enabling further spectral analysis and making it a forward-looking spectroscopic tool.

Surface-Enhanced Raman Spectroscopy for Monitoring the Biochemical Changes Due to DNA Mutations Induced by CRISPR-Cas9 Genome Editing in the Aspergillus niger Fungus.

In this study, surface-enhanced Raman spectroscopy (SERS) technique, along with principal component analysis (PCA) and partial least-squares discriminant analysis (PLS-DA), is used as a simple, quick, and cost-effective analysis method for identifying biochemical changes occurring due to induced mutations in the Aspergillus niger fungus strain. The goal of this study is to identify the biochemical changes in the mutated fungal cells (cell mass) as compared to the control/nonmutated cells. Furthermore, multivariate data analysis tools, including PCA and PLS-DA, are used to further confirm the differentiating SERS spectral features among fungal samples. The mutations are caused in A. niger by the clustered regularly interspaced palindromic repeat CRISPR-Cas9 genomic editing method to improve their biotechnological potential for the production of cellulase enzyme. SERS was employed to detect the changes in the cells of mutated A. niger fungal strains, including one mutant producing low levels of an enzyme and another mutant producing high levels of the enzyme as a result of mutation as compared with an unmutated fungal strain as a control sample. The distinctive features of SERS corresponding to nucleic acids and proteins appear at 546, 622, 655, 738, 802, 835, 959, 1025, 1157, 1245, 1331, 1398, and 1469 cm-1. Furthermore, PLS-DA is used to confirm the 89% accuracy, 87.7% precision, 87% sensitivity, and 88.9% specificity of this method, and the value of the area under the curve (AUROC) is 0.67. It has been shown that surface-enhanced Raman spectroscopy is an effective method for identifying and differentiating biochemical changes in genome-modified fungal samples.

Optical and scintillation properties of Er-doped La3Ga5SiO14 single crystals emitting near-infrared photons

0.5, 1, 2, and 5% Er-doped La3Ga5SiO14 (LGS) single crystals were synthesized by the Floating Zone method and the photoluminescence (PL) and scintillation properties were characterized. Emission peaks appeared at visible–near-infrared (NIR) regions. Their spectral features matched with those due to the 4f–4f transitions of Er3+. The PL quantum yields of 0.5, 1, 2, and 5% Er-doped samples were respectively 45.0, 73.3, 82.5, and 69.4%. The decay times due to the 2H9/2–4I15/2 transitions of Er3+ were obtained when visible ranges were monitored under excitation of both visible lights and X-rays. All the prepared samples showed good linearity between the X-ray exposure dose rate and scintillation signals in NIR ranges. The best lower detection limit in prepared samples was 5 mGy/h recorded by the 2% Er-doped LGS.

WPS:A whole phenology-based spectral feature selection method for mapping winter crop from time-series images

Accurately obtaining the spatial distribution and planting patterns of crops is very important for agricultural planning and food security. At present, time-series images have been proved to be an effective tool to characterize crop seasonal growth patterns, and identifying crop information by measuring the time-series similarity between unknown classes and known crop phenology curves is also considered to be a useful way. However, the existing methods of selecting feature ignore the connection between each phenological stage of crops and the unique growth rules of the whole phenology, which makes it difficult to select time-series spectral features that are potentially important for crop mapping. In order to make up for this problem, a Whole Phenology-based Spectral Feature Selection (WPS) method was proposed. The aim was to select the time-series feature sets with great differences among winter crops from a large number of spectral features, so as to improve the mapping accuracy of winter rapeseed and winter wheat. Firstly, spectral separability between all classes is calculated. Secondly, the key phenological periods of winter crops were selected according to the importance of temporal features, and the spectral feature sets with high separability were selected according to the key phenological periods. Finally, a Time-weighted Dynamic Time Warping (TWDTW) algorithm was used to generate the winter rapeseed and winter wheat maps of two cities in the middle and lower reaches of the Yangtze River. The mapping accuracy of the two crops is more than 92%, which matches the crop planting area well. The research shows that combining the WPS method with the TWDTW mapping method has great potential to accurately map crop types based on satellite time-series images.

Line graph contrastive learning for node classification

Existing graph contrastive learning methods often rely on differences in node features within subgraphs, lacking effective capture of the global structural information of the graph. To address this issue, we propose a novel graph contrastive learning method, Line Graph Contrastive Learning (LineGCL), aimed at overcoming the deficiencies in understanding the overall characteristics and topological structures of graphs found in current methods. LineGCL utilizes the characteristics of line graphs to transform the original graph into corresponding line graphs, presenting edge information in the form of node features, effectively capturing the global structural information of the graph. Additionally, LineGCL adopts a novel multi-view contrastive learning approach, characterizing the similarity and differences between the original graph and line graph comprehensively from the perspectives of node features and spectral features, further enhancing the model's understanding and learning capabilities of the global structure of graphs. Experimental results on public datasets demonstrate that LineGCL outperforms baseline models, achieving performance improvements ranging from 0.5% to 28.5%. These results validate the effectiveness and superiority of LineGCL in capturing and understanding global structural information, surpassing the limitations of baseline methods in graph contrastive learning.

Hyperspectral Image Classification Based on 3D–2D Hybrid Convolution and Graph Attention Mechanism

Convolutional neural networks and graph convolutional neural networks are two classical deep learning models that have been widely used in hyperspectral image classification tasks with remarkable achievements. However, hyperspectral image classification models based on graph convolutional neural networks using only shallow spectral or spatial features are insufficient to provide reliable similarity measures for constructing graph structures, limiting their classification performance. To address this problem, we propose a new end-to-end hyperspectral image classification model combining 3D–2D hybrid convolution and a graph attention mechanism (3D–2D-GAT). The model utilizes the collaborative work of hybrid convolutional feature extraction module and GAT module to improve classification accuracy. First, a 3D–2D hybrid convolutional network is constructed and used to quickly extract the discriminant deep spatial-spectral features of various ground objects in hyperspectral image. Then, the graph is built based on deep spatial-spectral features to enhance the feature representation ability. Finally, a network of graph attention mechanism is adopted to learn long-range spatial relationship and distinguish the intra-class variation and inter-class similarity among different samples. The experimental results on three datasets, Indian Pine, the University of Pavia and Salinas Valley show that the proposed method can achieve higher classification accuracy compared with other advanced methods.

The GAPS Programme at TNG. LV. Multiple molecular species in the atmosphere of HAT-P-11 b and review of the HAT-P-11 planetary system

The atmospheric characterisation of hot and warm Neptune-size exoplanets is challenging mainly due to their relatively small radius and atmospheric scale height, which reduce the amplitude of atmospheric spectral features. The warm-Neptune HAT-P-11\,b is a remarkable target for atmospheric characterisation because of the large brightness of its host star ($V=9.46$\,mag; $H=7.13$\,mag). The aims of this work are to review the main physical and architectural properties of the HAT-P-11 planetary system, and to probe the presence of eight molecular species in the atmosphere of HAT-P-11\,b through near-infrared (NIR) high-resolution transmission spectroscopy. We reviewed the physical and architectural properties of the HAT-P-11 planetary system by analysing transits and occultations of HAT-P-11\,b from the Kepler data set as well as HIRES at Keck archival radial-velocity data. We modelled the latter with Gaussian-process regression and a combined quasi-periodic and squared-exponential kernel to account for stellar variations on both (short-term) rotation and (long-term) activity-cycle timescales. In order to probe the atmospheric composition of HAT-P-11\,b, we observed four transits of this target with the NIR GIANO-B at TNG spectrograph and cross-correlated the data with template atmospheric transmission spectra. We find that the long-period radial-velocity signal previously attributed to the HAT-P-11\,c planet ($P 9.3$\,years; p i J $; $e 0.6$) is more likely due to the stellar magnetic activity cycle. Nonetheless, the Hipparcos-Gaia difference in the proper-motion anomaly suggests that an outer-bound companion might still exist. For HAT-P-11\,b, we measure a radius of p J $, a mass of $M_ p J $, a bulk density of $ p $, and an orbital eccentricity of $e=0.2577^ $. These values are compatible with those from the literature. Probing its atmosphere, we detect the presence of two molecular species H2O and NH3 with a S/N of $5.1$ and $5.3$, and a significance of $3.4\ and $5.0\ respectively. We also tentatively detect the presence of CO2 and CH4 with a S/N of $3.0$ and $4.8$, and a significance of $3.2\ and $2.6\ respectively. We revisit the HAT-P-11 planetary system, confirm the presence of water vapour, and report the detection of NH3 in the atmosphere of HAT-P-11\,b, also finding hints for the presence of CO2 and CH4 that need to be confirmed by further observations.