Convolutional neural networks (CNNs) have been widely used in hyperspectral image (HSI) classification tasks because of their excellent local spatial feature extraction capabilities. However, it is difficult to establish dependencies between long sequences data for CNNs, there are limitations the process processing spectral sequence features. To overcome these limitations, inspired by Transformer model, a spatial–spectral transformer with cross-attention (CASST) method proposed. Overall,...

In cross-domain hyperspectral image (HSI) classification, the labeled samples of target domain are very limited, and it is a worthy attention to obtain sufficient class information from source categorize classes (both same new unseen classes). This article investigates this problem by employing few-shot learning (FSL) in meta-learning paradigm. However, most existing FSL methods extract statistical features based on convolutional neural networks (CNNs), which typically only consider local...

Hyperspectral images (HSIs) contain abundant information in the spatial and spectral domains, allowing for a precise characterization of categories materials. Convolutional neural networks (CNNs) have achieved great success HSI classification, owing to their excellent ability local contextual modeling. However, CNNs suffer from fixed filter weights deep convolutional layers, which lead limited receptive field high computational burden. The recent Vision Transformer (ViT) models long-range...

Deep neural networks play a significant role in hyperspectral image (HSI) processing, yet they can be easily fooled when trained with adversarial samples (generated by adding tiny perturbations to clean samples). These are invisible the human eye, but lead misclassification deep learning model. Recent research on defense against HSI classification has improved robustness of exploiting global contextual information. However, available methods do not distinguish between different classes...

This letter introduces a novel spatial-spectral classification method for hyperspectral images (HSIs) based on structural-kernel collaborative representation (SKCR), which considers one weak assumption of spatial neighborhood that the pixels in superpixel belong to same class when exploiting contextual information HSI. The proposed consists following steps. First, segmentation strategy is used construct self-adaptive regions Then, structural within each block extracted density peak and K...

The multisource remote sensing classification task has two main challenges. 1) How to capture hyperspectral image (HSI) and light detection ranging (LiDAR) features cooperatively fully mine the complementary information between data. 2) adaptively fuse features, which should not only overcome imbalance HSI LiDAR data but also avoid generation of redundant information. local interaction transformer (LIIT) model proposed herein can effectively address these above issues. Specifically,...

Feature extraction is known to be an effective way in both reducing computational complexity and increasing accuracy hyperspectral imagery (HSI) classification. In this article, we propose a novel HSI feature based on joint adaptive structure density (JASD), which can make full use of the texture level effectively utilize spatial pixel level. Specifically, proposed JASD method considers two weak assumptions when exploiting information contained HSI: 1) close pixels spectral space have same...

Window-based operation is a general technique for hyperspectral anomaly detection. However, the problem remains that background knowledge containing abnormal information often affects attributes of test pixels. In this article, dual collaborative representation (DCR)-based detection method proposed to solve above effectively, which consists following main steps. First, low-rank and sparse matrix decomposition employed obtain matrix. Then, density peak clustering algorithm applied calculate...

Transporter engineering is an effective strategy for enhancing the transmembrane transfer of target substrates and alleviating feedback inhibition in microbial cells. The LysE transporter, a key indicator both L-arginine (L-Arg) L-lysine (L-Lys) secretion Corynebacterium glutamicum, plays crucial role efficient synthesis these amino acids. Owing to its broad substrate specificity, mutant with high specificity L-Arg extrusion essential achieving production. In this study, we constructed...

Spectral–spatial information plays an essential role in hyperspectral image (HSI) classification compared to pure spectral information. However, the neighbor spectral–spatial of a pixel tends be mixed into other ground coverings due various external factors such as weather and sensor jitter, mainstream HSI methods present low sensitivity for spatial this situation. This article proposes novel feature extraction method via 3-D block characteristics sharing (3-D-BCS) that redefines...

In hyperspectral image (HSI) classification, the performance of supervised learning tends to be affected by priori knowledge, i.e., quantity and quality samples. However, it is inevitable limit classification due presence noisy labels in training this article, we first propose a hierarchical constrained energy minimum (HCEM) method detect mislabeled samples (noisy labels) original set trained with task boost classifiers spectral-spatial methods HSI applications. The basic idea behind work...

In recent times, multisource remote sensing technology [e.g., hyperspectral image (HSI) and light detection ranging (LiDAR) data] has been widely used in urban land-use recognition owing to its high classification effectiveness compared using only single-source data. this study, a multiview hierarchical network (MVHN) technique is developed for HSI LiDAR data classification, which conducts the following execution procedures. First, based on preset band step length, original sampled divided...

Steel surface demonstrates various sorts of defects due to the production technique and environment. The appearance defect is in much more random pattern than that normal texture image. Therefore, it challenging capture discriminant information categorize defects. image out registration grayscale, thus, local descriptor inclined be utilized for feature extraction. In previous works, involving a categorizing images, thresholding operator participates hand-crafted extraction, such as binary...

This work adopts sparse coding and spatial pyramid matching to classify the vehicle images. The targets of interest, vehicles in images, are always degraded complex circumstance. Hence, it seems difficult carry out classification task by methods combined gray feature traditional classifiers. Considering image without assignment influence caused weather, this paper proposes a method based on matching. First, proposed extracts patch-based computed with discriminate dictionary. With dualizing...

Convolutional neural network (CNN) is widely used in HSI classification owing to their advantages of spatial-spectral features capture capability, learning depth as well structural flexibility. Nevertheless, the shape convolution kernel fixed, a limitation that leads fixation when modeling different feature CNN, especially edge regions between classes. A multiscale adaptive (MSAC) model proposed this paper overcome shortcoming. Combining superpixels with traditional convolutional kernels...

Label information plays an important role in supervised high-resolution remote sensing (HRRS) image scene classification. However, the labels of a dataset are probably unreliable and may contain "noisy" labels. Focusing on uncertain problem, covariance matrix representation-based noisy label model (CMR-NLD) is designed for HRRS The main steps as follows. First, pretrained convolutional neural network employed to extract images deep features principal component analysis based dimensionality...

Deep convolutional neural networks play an important role in hyperspectral image (HSI) classification tasks through hierarchical learning. Recent work based on deep learning has made great progress exploring contextual features, with more of these approaches focusing nonlocal information. Nevertheless, the information obtained by methods still room for improvement as they only consider semantic level. Moreover, ignore importance features spectral domain, component especially HSIs. This...

With the high proportion accession of renewable energy, uncertainty power system gradually increases. Scenario generation is an important method to describe a system, and plays role in operation planning scheduling systems. In this work, we proposed wind photovoltaic random scenario based on Conditional Generative Adversarial Networks (CGAN), Combined with k-means clustering cluster label historical output data, mapping relationship between noise input real data under constraint learned...

Limited alliinase resources cause difficulties in the biosynthesis of thiosulfinates (e.g., allicin), restricting their applications agricultural and food industries. To effectively biosynthesize thiosulfinates, this study aimed to excavate bacterial elucidate catalytic properties. Two cystathionine β-lyases (MetCs) possessing high activity (>60 U mg

For displaying high-dynamic-range images acquired by thermal camera systems, 14-bit raw infrared data should map into 8-bit gray values. This paper presents a new method for detail enhancement of to display the image with relatively satisfied contrast and brightness, rich information, no artifacts caused processing. We first adopt propagated filter smooth input separate base layer layer. Then, we refine using modified histogram projection compressing. Meanwhile, adaptive weights derived from...