A5101678568- Particle physics theoretical and experimental studies
- Quantum Chromodynamics and Particle Interactions
- High-Energy Particle Collisions Research
- Particle Detector Development and Performance
- Bayesian Methods and Mixture Models
- Statistical Methods and Inference
- Advanced Image and Video Retrieval Techniques
- Medical Image Segmentation Techniques
- Image and Signal Denoising Methods
- Image Retrieval and Classification Techniques
- Traffic Prediction and Management Techniques
- Robotics and Sensor-Based Localization
- Radiomics and Machine Learning in Medical Imaging
- Image Processing Techniques and Applications
- Statistical Methods and Bayesian Inference
- Soil Geostatistics and Mapping
- Image Processing and 3D Reconstruction
- Explainable Artificial Intelligence (XAI)
- Multimodal Machine Learning Applications
- Remote Sensing and Land Use
- Medical Imaging and Analysis
- Theoretical and Computational Physics
- Blind Source Separation Techniques
- Stochastic processes and statistical mechanics
- Advanced Vision and Imaging
Shandong University
2025
First Affiliated Hospital of Nanchang University
2024
Northeastern University
2018-2024
Northwest A&F University
2024
Nanchang University
2024
Iowa State University
2024
Nanjing University of Aeronautics and Astronautics
2024
University of Science and Technology Beijing
2023
University of Electronic Science and Technology of China
2008-2023
Chongqing Three Gorges University
2010-2023
Parameter estimation from noisy versions of realizations Markov models is extremely difficult in all but very simple examples. The paper identifies these difficulties, reviews ways coping with them practice, and discusses detail a class methods Monte Carlo flavour. Their performance on examples suggests that they should be valuable, practically feasible procedures the context range otherwise intractable problems. An illustration provided based satellite data.
The creep camber of simply supported beam bridges in high-speed railways affects ride comfort and threatens driving safety, necessitating accurate prediction. This study monitors environmental parameters to collect detection data for prediction purposes. interquartile range (IQR) method is employed process outliers the raw data, simple wavelet transform applied noise reduction. To incorporate temporal information, time information converted into features such as solar incidence angles...
Abstract Chinese jujube (Ziziphus jujuba Mill.) is one of the most important deciduous tree fruits in China, with substantial economic and nutritional value. Jujube was domesticated from its wild progenitor, (Z. var. spinosa), both have high medicinal Here we report 767.81- 759.24-Mb haplotype-resolved assemblies a dry-eating ‘Junzao’ (JZ) accession (SZ), using combination multiple sequencing strategies. Each assembly yielded two complete genomes at telomere-to-telomere (T2T) level, ~81.60...
Based on deformable templates, the paper formulates an integrated and flexible Bayesian recognition system of multiple occluded objects. Various local dependence properties model are obtained to reduce computational cost with increase in number Numerical results for a synthetic image real mushrooms discussed.
The paper investigates parameter estimation for Markov random fields and hidden fields, where noisy data are available. EM algorithms described an approximate procedure is developed based on row-by-row relaxation analysis. Numerical illustrations provided.
Background: In recent years, computer-aided diagnosis (CAD) systems have played an important role in breast cancer screening and diagnosis. The image segmentation task is the key step a CAD system for rapid identification of lesions. Therefore, efficient network necessary improving diagnostic accuracy screening. However, due to characteristics blurred boundaries, low contrast, speckle noise ultrasound images, lesion challenging. addition, many proposed tumor networks are too complex be...
Despite the recent progress in deep neural networks (DNNs), it remains challenging to explain predictions made by DNNs. Existing explanation methods for DNNs mainly focus on post-hoc explanations where another explanatory model is employed provide explanations. The fact that can fail reveal actual original reasoning process of raises need build with built-in interpretability. Motivated this, many self-explaining have been proposed generate not only accurate but also clear and intuitive...
Interpreting deep neural networks through examining neurons offers distinct advantages when it comes to exploring the inner workings of Deep Neural Networks. Previous research has indicated that specific within vision possess semantic meaning and play pivotal roles in model performance. Nonetheless, current methods for generating neuron semantics heavily rely on human intervention, which hampers their scalability applicability. To address this limitation, paper proposes a novel post-hoc...
This paper presents a simulation study of block (one line or two lines pixels) updating for Markov random fields. Point and relaxation methods are compared. Some pseudo-likelihoods, based on the conditional density pixels, used together with modified EM algorithms to estimate parameters from noisy images.
Hyperspectral images classification relies on the accurate and efficient extraction of discriminative features, detail preservation, learning with limited training samples. This article, therefore, presents an advanced neural network architecture combined convolutional conditional random fields (ConvCRF) region growing (RGW) approaches to address these key issues. First, a depthwise separable fully residual (DFRes) is proposed for feature learning, where operation ensures larger field view,...
Route travel time varies with vehicles and traffic demand. Besides the average route time, reliability in form of distribution is indispensable. However, sample size Complete Travel Times (TTC) rather small for many reasons. Existing methods using convolution rely on strong assumptions about correlation structure or link distributions; other relying scaled Partial (TTP) may extend estimation bias. To overcome these issues, we present an method by fusing kinds information from Automatic...
Abstract Methods are presented for detecting ridges and/or antiridges using noisy data. Several alternative criteria proposed identifying points on a ridge and procedures following line discussed. The methods illustrated by examples.
The <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k\text {-means}$ </tex-math></inline-formula> clustering problem concerns finding a partition of the data points into notation="LaTeX">$k$ clusters such that total within-cluster squared distance is minimized. This optimization objective non-convex, and not everywhere differentiable. In general, there exist spurious local solutions other than global...
Video stabilization is a technology to remove the dithering between frames in video sequence by motion estimation, filter and compensation. In this paper, we focus on estimation (ME) step which known as most important part for successful system because latter processing heavily relies it. However, traditional ME methods mainly hypothesize that movement within only consists of displacement. So their will lose validity occasion when rotation also presented. To solve problem, interest points...
The Expectation-Maximization algorithm is perhaps the most broadly used for inference of latent variable problems. A theoretical understanding its performance, however, largely remains lacking. Recent results established that EM enjoys global convergence Gaussian Mixture Models. For Mixed Linear Regression, only local have been established, and those high SNR regime. We show here converges mixed linear regression with two components (it known it may fail to converge three or more), moreover...
Inter-regional energy dispatch and regional peak cutting valley filling require accurate load forecasting as support. In order to improve the accuracy, this paper proposes a multi-model fusion method based on CNN (convolutional neural network)-LSTM (long short-term memory)-LGBM (Light Gradient Boosting Machine) considering demand response. The CNN's ability is exploited effectively extract local features, LSTM's grasp time series information used build serial CNN-LSTM model. Meanwhile,...
Purpose. To build a correction factor dataset to improve accuracy of dose calculations for spot-scanning proton therapy with range shifter employed. Material and method. Our synchrotron-based spot scanning system has an energy 72.5–221.8 MeV. The Bragg peak chamber was used perform the integral depth measurements in water phantom different air-gap sizes between phantom. Three typical energies 151.0, 181.1 219.3 MeV three 10, 20 30 cm were selected measurements. measured data as standard...