A5100734649- Remote-Sensing Image Classification
- Advanced Image Fusion Techniques
- Image and Signal Denoising Methods
- Remote Sensing and Land Use
- Infrared Target Detection Methodologies
- Adaptive Control of Nonlinear Systems
- Face and Expression Recognition
- Advanced Image Processing Techniques
- Advancements in Battery Materials
- Machine Learning and ELM
- Advanced Image and Video Retrieval Techniques
- Advanced Battery Materials and Technologies
- Photoacoustic and Ultrasonic Imaging
- Advanced Neural Network Applications
- Geometric Analysis and Curvature Flows
- Domain Adaptation and Few-Shot Learning
- Spectroscopy and Chemometric Analyses
- Image Retrieval and Classification Techniques
- Robotics and Sensor-Based Localization
- Robot Manipulation and Learning
- Stability and Control of Uncertain Systems
- Virtual Reality Applications and Impacts
- Computational Fluid Dynamics and Aerodynamics
- Supercapacitor Materials and Fabrication
- Advanced Battery Technologies Research
Xidian University
2016-2025
Shandong Xiehe University
2025
Fudan University
2025
The University of Sydney
2025
Hefei University
2025
Beihang University
2024
Beijing Institute of Technology
2024
Wuhan University
2017-2024
Qufu Normal University
2022-2024
Nanjing University of Information Science and Technology
2024
Abstract Pursuit of advanced batteries with high‐energy density is one the eternal goals for electrochemists. Over past decades, lithium–sulfur (LSBs) have gained world‐wide popularity due to their high theoretical energy and cost effectiveness. However, road market still full thorns. Apart from poor electronic conductivity sulfur‐based cathodes, LSBs involve special multielectron reaction mechanisms associated active soluble lithium polysulfides intermediates. Accordingly, electrode design...
This paper reviews the second challenge on spectral reconstruction from RGB images, i.e., recovery of whole- scene hyperspectral (HS) information a 3-channel image. As in previous challenge, two tracks were provided: (i) "Clean" track where HS images are estimated noise-free RGBs, themselves calculated numerically using ground-truth and supplied sensitivity functions (ii) "Real World" track, simulating capture by an uncalibrated unknown camera, recovered noisy JPEG-compressed images. A new,...
Recently, embedding and metric-based few-shot learning (FSL) has been introduced into hyperspectral image classification (HSIC) achieved impressive progress. To further enhance the performance with few labeled samples, we in this paper propose a novel FSL framework for HSIC class-covariance metric (CMFSL). Overall, CMFSL learns global class representations each training episode by interactively using samples from base classes, synthesis strategy is employed on classes to avoid overfitting....
With success of convolutional neural networks (CNNs) in computer vision, the CNN has attracted great attention hyperspectral classification. Many deep learning-based algorithms have been focused on feature extraction for classification improvement. In this letter, a novel learning framework based fully is proposed. Through convolution, deconvolution, and pooling layers, features data are enhanced. After enhancement, optimized extreme machine (ELM) utilized The proposed outperforms existing...
Recent promising effort for spectral reconstruction (SR) focuses on learning a complicated mapping through using deeper and wider convolutional neural networks (C- NNs). Nevertheless, most CNN-based SR algorithms neglect to explore the camera sensitivity (CSS) prior interdependencies among intermediate features, thus limiting representation ability of network performance SR. To conquer these issues, we propose novel adaptive weighted attention (AWAN) SR, whose backbone is stacked with...
Convolutional neural networks (CNNs) have recently achieved impressive improvements on hyperspectral (HS) pansharpening. However, most of the CNN-based HS pansharpening approaches would to first upsample low-resolution image (LR-HSI) using bicubic interpolation or data-driven training strategy, which inevitably lose some details greatly rely learning process. In addition, previous methods regard as a black-box problem and treat diverse features equally, thus hindering discriminative ability...
With success of Deep Belief Networks (DBNs) in computer vision, DBN has attracted great attention hyperspectral classification. Many deep learning based algorithms have been focused on feature extraction for classification improvement. Multi-features, such as texture feature, are widely utilized process to enhance accuracy greatly. In this paper, a novel framework an optimal and enhancement (TFE) is proposed. Through band grouping, sample selection guided filtering, the features data...
Hyperspectral image (HSI) super-resolution (SR) is a challenging task due to the problems of texture blur and spectral distortion when upscaling factor large. To meet these two challenges, band attention through adversarial learning method proposed in this article. First, we put SR process generative network (GAN) framework, so that resulted high-resolution HSI can keep more details. Second, different from other band-by-band method, input our full bands. In order explore correlation bands...
An integral imaging-based 2D/3D convertible display system is proposed by using a lens-array holographic optical element (LAHOE), polymer dispersed liquid crystal (PDLC) film, and projector. The LAHOE closely attached to the PDLC film constitute projection screen. used realize imaging 3D display. When with an applied voltage in transparent state, projector projects Bragg matched image, works mode. without scattering 2D A prototype of developed, it provides images properly.
In this article, we propose a weakly supervised low-rank representation (WSLRR) method for hyperspectral anomaly detection (HAD), which formulates deep learning-based HAD into low-lank optimization problem not only characterizing the complex and diverse background in real HSIs but also obtaining relatively strong supervision information. Different from existing unsupervised methods, first model manner, achieves better performance without prior information is restrained by richly correct...
Hyperspectral imagery with very high spectral resolution provides a new insight for subtle nuances identification of similar substances. However, hyperspectral target detection faces significant challenges intraclass dissimilarity and interclass similarity due to the unavoidable interference caused by atmosphere, illumination, sensor noise. In order effectively alleviate these inconsistencies, this paper proposes novel method without strict assumptions on data distribution based an...
This paper reviews the third biennial challenge on spectral reconstruction from RGB images, i.e., recovery of whole-scene hyperspectral (HS) information a 3-channel image. presents "ARAD_1K" data set: new, larger-than-ever natural image set containing 1,000 images. Challenge participants were required to recover hyper-spectral synthetically generated JPEG-compressed images simulating capture by known calibrated camera, operating under partially parameters, in setting which includes...
In practice, the acquirement of labeled samples for hyperspectral image (HSI) is time-consuming and labor-intensive. It frequently induces trouble model overfitting performance degradation supervised methodologies in HSI classification (HSIC). Fortunately, semisupervised learning can alleviate this deficiency, graph convolutional network (GCN) one most effective approaches, which propagates node information from each other a transductive manner. study, we propose cross-scale prototypical...
Semantic segmentation for remote sensing images (RSI) is critical the Earth monitoring system. However, covariate shift between RSI datasets under different capture conditions cannot be alleviated by directly using unsupervised domain adaptation (UDA) method, which negatively affects accuracy in RSI. We propose a stepwise adaptive network with alleviation (Cov-DA) parsing to solve this issue. Specifically, alleviate generated sensors, both source and target domains are projected into...
Cross-domain (CD) hyperspectral image classification (HSIC) has been significantly boosted by methods employing Few-Shot Learning (FSL) based on CNNs or GCNs. Nevertheless, the majority of current approaches disregard prior information spectral coordinates with limited interpretability, leading to inadequate robustness and knowledge transfer. In this paper, we propose an asymmetric encoder-decoder architecture, Spectral Coordinate Transformer (SCFormer), for CDFSL HSIC task. Several dense...
Remote sensing object detection (RSOD) is a fundamental and valuable task in Earth monitoring. However, remote images (RSIs) are typically acquired from bird's eye perspective, resulting intrinsic properties such as the complex backgrounds, random dense distribution of objects, multiscale objects. These hinder direct application well-performed methods natural (NIs) domain to RSIs domain, thereby limiting attainment desired performance. To address this, we propose pyramid convolutional vision...
<title>Abstract</title> <bold>IMPORTANCE</bold> Understanding trends in cardiovascular and diabetes-related metabolic biomarkers across populations, especially during the COVID-19 pandemic, is essential for informing public health strategies targeting prevention management of diseases (CVD) diabetes. This study aimed to assess among U.S. adults from 2013-2014 2021-2023. <bold>DESIGN, SETTING, AND PARTICIPANTS</bold> analyzed five cycles cross-sectional data National Health Nutrition...
Conventional classification algorithms have shown great success for balanced classes. In remote sensing applications, it is often the case that classes are imbalanced. This paper proposes a novel solution to solve problem of imbalanced training samples in hyperspectral image classification. It consists two parts: one large-size sample sets and other small-size sets. Specifically, an algorithm based on orthogonal complement subspace projection (OCSP) proposed select from classes, also OCSP...
Hyperspectral (HS) pansharpening, as a special case of the superresolution (SR) problem, is to obtain high-resolution (HR) image from fusion an HR panchromatic (PAN) and low-resolution (LR) HS image. Though pansharpening based on deep learning has gained rapid development in recent years, it still challenging task because following requirements: 1) unique model with goal fusing two images different dimensions should enhance spatial resolution while preserving spectral information; 2) all...
Triple negative breast cancer (TNBC) is the most aggressive subtype of cancer, with ineffective treatment and poor prognosis. It in great demand to develop a novel theranostic strategy for accurate diagnosis targeted TNBC. In present study, one nanoplatform (HA-ICG-Fe-PDA), endowed multimodal imaging-guided chemodynamic/photodynamic/photothermal (CDT/PDT/PTT) synergistic therapy capacity toward TNBC, was innovatively constructed. The prepared by covalently conjugating ICG-decorated...