Article
Computer Science, Information Systems
Xingyu Fu, Bin Fang, Mingliang Zhou, Sam Kwong
Summary: This paper presents a new method for image segmentation, using adaptively weighted signed pressure force and Legendre polynomial method to drive an active contour, ensuring high accuracy and computational efficiency for images with inhomogeneous intensity, blurred edges, low contrast, and noise problems.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Yunyun Yang, Ruicheng Xie, Wenjing Jia, Gang Zhao
Summary: This paper introduces a double level set segmentation model based on mutual exclusion, which accurately and independently segments adjacent regions while maintaining their independence. Experimental results demonstrate the model's high accuracy in segmenting adjacent tissues in brain, as well as its robustness to intensity inhomogeneity and noise in synthetic images. Comparisons with other models show that the double level sets model outperforms classical models in segmenting adjacent tissues.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Information Systems
Jiajie Zhu, Bin Fang, Mingliang Zhou, Futing Luo, Weizhi Xian, Gang Wang
Summary: This paper proposes an active contour model based on adaptively variable exponent combining Legendre polynomial for image segmentation. By defining Legendre polynomial intensity, adaptively LPI term, and introducing distance regularization term, the method demonstrates strong robustness and adaptability in handling intensity inhomogeneity and blurred boundaries.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Engineering, Electrical & Electronic
Ming Dai, Zhiheng Zhou, Tianlei Wang, Yongfan Guo
Summary: In this paper, a novel segmentation model using generalized divergences is proposed based on the traditional level set method. The main advantage of generalized divergences is their smooth connection performance in measuring the discrepancy between two probability distributions of segmented image parts. Experimental results demonstrate the superior performance of the proposed method in both qualitative and quantitative aspects compared to previous active contour models.
CIRCUITS SYSTEMS AND SIGNAL PROCESSING
(2021)
Article
Engineering, Electrical & Electronic
Bijay Kumar Sa, Rutuparna Panda, Sanjay Agrawal
Summary: In this study, a new adaptively weighted level-set evolution method based on relevant edge probability is investigated for medical image segmentation. By adjusting the weights according to the image's relative value, the leakage and premature convergence are reduced, leading to improved segmentation accuracy.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Computer Science, Information Systems
Amirhossein Fallahdizcheh, Sandeep Laroia, Chao Wang
Summary: In this article, a two-stage active contour method is proposed to accurately and efficiently segment ascites. A morphological-driven thresholding method is used to automatically locate the initial contour of the ascites, and a novel sequential active contour algorithm is applied to accurately segment the ascites from the background. The experimental results show the superiority of the proposed method in both accuracy and time efficiency.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Optics
Yang Chen, Guirong Weng
Summary: This research proposes an active contour model based on local pre-piecewise fitting image, which improves robustness by optimizing the gradient descent flow formula and regularization term. It effectively segments images with uneven grayscale and has obvious advantages in initialization and segmentation efficiency.
Article
Computer Science, Information Systems
Daniel Reska, Marek Kretowski
Summary: This paper presents a fast multi-stage image segmentation method that incorporates texture analysis into a level set-based active contour framework. It allows integrating multiple feature extraction methods without the need for prior knowledge of image patterns, and achieves high performance through GPU acceleration. The method is validated on synthetic and natural images and compared with similar algorithms.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Jiangxiong Fang, Huaxiang Liu, Jun Liu, Haiying Zhou, Liting Zhang, Hesheng Liu
Summary: This paper introduces a novel global and local fuzzy image fitting (GLFIF) based active contour model for image segmentation. By designing global and local fitted images and constructing an energy function, it addresses intensity inhomogeneity in images and proves convexity to ensure segmentation independence of initialization. Experimental results demonstrate the model's robustness in segmenting images.
APPLIED SOFT COMPUTING
(2021)
Article
Engineering, Multidisciplinary
Xin Yan, Guirong Weng
Summary: The paper proposes a hybrid active contour model driven by optimized local pre-fitting image energy for fast image segmentation, which effectively handles images with intensity inhomogeneity and noise interference. By combining different pre-fitting functions and an optimized edge indicator function, the proposed model shows high efficiency and robustness in initializing parameters and segmenting images with various characteristics.
APPLIED MATHEMATICAL MODELLING
(2022)
Article
Engineering, Electrical & Electronic
Junwei Li, Peng Jiang, He Zhu
Summary: A new active contour model for image segmentation is proposed in this article, which includes local texture information and a Bayesian framework to counteract noise and boundary pollution, improving segmentation accuracy.
IEEE SENSORS JOURNAL
(2021)
Article
Multidisciplinary Sciences
Mei Zhang, Dan Meng, Lingling Liu, Jinghua Wen
Summary: This paper proposes an improved algorithm based on the no-weight initialization level set model to address the shortcomings of the traditional level set model. The improved method introduces bilateral filters and uses implicit surface level sets to accurately extract and segment the original target image object. Experimental results demonstrate that the improved method achieves better edge contour extraction and noise reduction compared to the traditional non-reinitialized level set model.
Article
Computer Science, Artificial Intelligence
Elizangela de Souza Reboucas, Fatima Nelsizeuma Sombra de Medeiros, Regis Cristiano P. Marques, Joao Victor S. Chagas, Matheus T. Guimaraes, Lucas O. Santos, Aldisio G. Medeiros, Solon A. Peixoto
Summary: Scientific research on methodologies and algorithms to enhance medical diagnostic support remains a top priority, with computer-aided diagnostic systems utilizing the IoT showing promise in improving accessibility and integration. The proposed FLog Parzen Level Set method achieved stable and satisfactory results with low computational costs, demonstrating high accuracy, sensitivity, and MCC values across various datasets.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Bin Dong, Guirong Weng, Ri Jin
Summary: An unsupervised active contour model with Self Organizing Maps (SOM) is proposed in this paper, which effectively segments images with intensity inhomogeneity.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Information Systems
Shimaa Fawzy, Hossam El-Din Moustafa, Ehab H. H. AbdelHay, Mohamed Maher Ata
Summary: The presented work proposes a robust approach for skin lesion segmentation using optimized active contours and level sets. The proposed system accurately separates the region of interest by defining the appropriate contour or curvature. The results show that the proposed Sine-IPO algorithm achieves the best segmentation performance for different types of skin lesions in dermoscopy images.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Software Engineering
Lindong Li, Linbo Qing, Yuchen Wang, Jie Su, Yongqiang Cheng, Yonghong Peng
Summary: Social relations and interactions are fundamental to human society, and effective recognition of these relationships can greatly impact understanding and improving people's psychology and behaviors. This paper introduces a new framework, HF-SRGR, which utilizes a hybrid feature extraction and fusion approach to recognize social relations. Through constructing a social relation graph, designing an attention mechanism, and enhancing scene information propagation, the proposed method achieves better accuracy compared to existing state-of-the-art methods.
Article
Computer Science, Information Systems
Shuyan Huang, Zhijing Yang, Yukai Shi, Junpeng Tan, Hao Li, Yongqiang Cheng
Summary: This paper proposes two transformer-based point-cloud-completion networks and a coarse-to-fine strategy to extract object shape features. It addresses the challenges of acquiring high-fidelity 3D models from real-world scans. Experimental results demonstrate the effectiveness of this method.
Article
Chemistry, Analytical
Yuchen Wang, Linbo Qing, Zhengyong Wang, Yongqiang Cheng, Yonghong Peng
Summary: Social relationships are important for understanding human behavior and developing complex social intelligent systems. This study proposes a new framework for social relationship recognition that effectively integrates features on different scales and employs attention mechanisms to improve relationship representations. Experimental results demonstrate the effectiveness of the framework, including addressing data imbalance challenges.
Article
Energy & Fuels
Lu Zhang, Siyue Lu, Yifeng Ding, Dapeng Duan, Yansong Wang, Peiyi Wang, Lei Yang, Haohao Fan, Yongqiang Cheng
Summary: This paper aims to establish a short-term load probability forecasting model for individual users using smart meter data. The proposed RF-KDE method utilizes historical data and random forest to train a deterministic prediction model, and then employs kernel density estimation to obtain probability prediction results. The method shows promising results on a public dataset and offers advantages in parameter adjustment and training speed.
Article
Energy & Fuels
Xianglong Li, Longfei Ma, Ping Chen, Hui Xu, Qijing Xing, Jiahui Yan, Siyue Lu, Haohao Fan, Lei Yang, Yongqiang Cheng
Summary: The paper presents a probabilistic prediction model of solar irradiance based on XGBoost, which utilizes historical data to train a point prediction model and generates probability prediction intervals under different confidence levels using kernel density estimation. Experimental results demonstrate that this method has better accuracy and is suitable for engineering practice.
Article
Materials Science, Multidisciplinary
John A. Schneeloch, Yu Tao, Yongqiang Cheng, Luke Daemen, Guangyong Xu, Qiang Zhang, Despina Louca
Summary: Gapless Dirac magnons are present in bulk CrCl3, with inelastic neutron scattering intensity approaching zero at low temperatures. Upon warming, magnon-magnon interactions induce strong renormalization and decreased lifetimes.
NPJ QUANTUM MATERIALS
(2022)
Article
Energy & Fuels
Bo Yang, Yongqiang Cheng, Kai Chen, Zhong Wei, Zhigang Lei, Guoxuan Li
Summary: The double carbon goal sets new requirements for the low-carbon development of industrial parks. This study focuses on improving the distillation process performance for ester hydrolysis to alcohol by combining reactive and extractive distillation with ionic liquids-based mixed solvents. Traditional organic solvents are screened based on relative volatility, and ionic liquid entrainers are selected using the COSMO-RS model. The results show that the new process has advantages in terms of energy, economic, and environmental performance.
Article
Energy & Fuels
Bernadette R. Cladek, A. J. Ramirez-Cuesta, S. Michelle Everett, Marshall T. McDonnell, Luke Daemen, Yongqiang Cheng, Paulo H. B. Brant Carvalho, Christopher Tulk, Matthew G. Tucker, David J. Keffer, Claudia J. Rawn
Summary: Natural hydrate deposits are a rich source of CH4. Recent studies have shown that CH4 can be extracted from hydrates by CO2 exchange, which also has the potential for carbon sequestration. Understanding the impact of guest variation in CH4, CO2, and mixed hydrates on their stability and the processes of methane extraction and CO2 sequestration is crucial. The use of inelastic neutron scattering has revealed the dynamic modes in hydrate structures and the behavior of CH4, providing valuable insights for further research.
Article
Engineering, Chemical
Yongqiang Cheng, Bo Yang, Guoxuan Li, Kai Chen, Zhong Wei, Xin Gao, Hong Li, Zhigang Lei
Summary: This study developed a reactive distillation coupled with extractive distillation process using ionic liquids as entrainers for the transesterification of n-prop-anol and methyl acetate. By using the COSMO-RS model, a suitable entrainer, 1-ethyl-3-methylimidazolium acetate, was selected. The process performance was significantly improved due to the strong hydrogen bond between the entrainer and the components. The new process showed higher energy efficiency and lower economic cost compared to the conventional process.
SEPARATION AND PURIFICATION TECHNOLOGY
(2022)
Article
Chemistry, Multidisciplinary
Nawei Shi, Huazhang Wang, Yongqiang Cheng
Summary: This paper proposes a Chinese reading comprehension algorithm, called the AT-CRF Reader, to address the challenges faced by traditional machine reading comprehension models such as information loss, lack of capability to retain long-distance dependence, and inability to handle unanswerable questions. Experimental results show that the AT-CRF Reader model achieves significant improvements in performance compared to the baseline model on two Chinese machine reading comprehension datasets.
APPLIED SCIENCES-BASEL
(2022)
Article
Automation & Control Systems
Xintao Ding, Yongqiang Cheng, Yonglong Luo, Qingde Li, Prosanta Gope
Summary: This article proposes a consensus defense method (Cons-Def) to defend against adversarial attacks. Cons-Def generates augmented examples based on intensity exchange on the input image components and implements classification and detection based on the consensus of the classifications. Experimental results show that Cons-Def is robust against both white-box and black-box adversarial attacks.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Review
Chemistry, Multidisciplinary
Xinbin Yu, Yongqiang Cheng, Yuanyuan Li, Felipe Polo-Garzon, Jue Liu, Eugene Mamontov, Meijun Li, David Lennon, Stewart F. Parker, Anibal J. Ramirez-Cuesta, Zili Wu
Summary: Understanding the structure-catalysis relationship in heterogeneous catalysis relies on spectroscopic and scattering tools, with neutron scattering providing unique information due to its interaction with nuclei. Neutron techniques, including vibrational spectroscopy, diffraction, and quasielastic scattering, have been extensively used to study surface adsorbates, reaction mechanisms, and catalyst structures. This review comprehensively summarizes recent advances in neutron scattering investigations of heterogeneous catalysis and discusses the challenges and future opportunities in this field.
Proceedings Paper
Nuclear Science & Technology
Chris W. Chapman, Goran Arbanas, Jesse Brown, Kemal Ramic, Yongqiang Cheng, Jiao Lin, Douglas L. Abernathy, Alexander I. Kolesnikov, Matthew B. Stone, Luke Daemen, Anibal Ramirez Cuesta, Xunxiang Hu
Summary: With the increasing interest in thermal neutron scattering data, there is a need for new experimental data to evaluate new or previously evaluated materials. This evaluation process involves three steps: computing the phonon characteristics, calculating the dynamic structure factor, and simulating the experimental data. The evaluation can be achieved using the code systems OCLIMAX and MCViNE, developed by instrument scientists at SNS.
15TH INTERNATIONAL CONFERENCE ON NUCLEAR DATA FOR SCIENCE AND TECHNOLOGY, ND2022
(2023)
Article
Materials Science, Multidisciplinary
Yu Tao, Luke Daemen, Yongqiang Cheng, Joerg C. Neuefeind, Despina Louca
Summary: Neutron scattering and first-principles calculations were used to study topological quantum magnets FeSn and Fe3Sn2. Both materials are metallic and have dispersionless flat bands with Dirac nodes at the K point. The localized structure of both compounds was not observed, indicating their metallic nature. In FeSn, the lattice constant in the c-axis decreased anomalously along with changes in the phonon spectra, suggesting the presence of magnetoelastic coupling and spin-phonon interactions. In contrast, Fe3Sn2 did not show any lattice anomaly, and the inelastic signal was mostly due to phonons.
Article
Chemistry, Multidisciplinary
Steven Balding, Amadou Gning, Yongqiang Cheng, Jamshed Iqbal
Summary: Robotic agents are now common in various environments and can undertake increasingly complex tasks. Collaborative robotics can leverage the capabilities of multiple agents to achieve more complex tasks and compensate for lacking systems in less sophisticated agents. The key challenge is enabling communication and environment mapping among the agents, considering the heterogeneity of sensors. The voxel grid approach is presented as a solution for decentralized robotic colonies, with comparisons between single-agent and multi-agent systems provided for validation.
APPLIED SCIENCES-BASEL
(2023)