Article
Computer Science, Information Systems
Xinyan Zhang, Peng Gao, Sunxiangyu Liu, Kongya Zhao, Guitao Li, Liuguo Yin, Chang Wen Chen
Summary: The study introduces a global-local adjusting dense super-resolution network (GLADSR) to improve image super-resolution performance while reducing parameters and computational cost, using a global-local adjusting module (GLAM) and a separable pyramid upsampling (SPU) module. Extensive experiments demonstrate the superiority of GLADSR over state-of-the-art methods with less parameters and lower computational cost.
IEEE TRANSACTIONS ON MULTIMEDIA
(2021)
Article
Neurosciences
Leonie Henschel, David Kuegler, Martin Reuter
Summary: This study fills the gap in existing methods in the field of high-resolution MRI (HiRes) by proposing a Voxel-size Independent Neural Network (VINN) for resolution-independent image segmentation tasks. The FastSurferVINN method achieves whole brain segmentation within the resolution range of 0.7-1.0 mm and significantly outperforms existing methods at various resolutions. Additionally, this method addresses the data imbalance problem in HiRes datasets and has important application prospects.
Article
Microbiology
Lasse Sprankel, Margot P. Scheffer, Sina Manger, Utz H. Ermel, Achilleas S. Frangakis
Summary: The nap particle is an immunogenic surface adhesion complex that plays a crucial role in cell motility and binding of bacteria to host cells. Cryo-electron tomography structures reveal the bound and released states of the nap particle, providing insights into the mechanism of bacterial adhesion.
Article
Physics, Fluids & Plasmas
Xiufeng Yang, Song-Charng Kong, Qingquan Liu
Summary: The SPH-ASR method for multiphase flows is improved by introducing a particle shifting technique to improve the distribution of particles, which considers the variable smoothing length. Additionally, the algorithm for adaptive resolution is optimized and extended to three-dimensional applications, showing significant reduction in computational cost with maintained accuracy.
Article
Geochemistry & Geophysics
Yongchao Zhang, Jiawei Luo, Yongwei Zhang, Yulin Huang, Xiaochun Cai, Jianyu Yang, Deqing Mao, Jie Li, Xingyu Tuo, Yin Zhang
Summary: This article introduces a low-complexity super-resolution technique based on adaptive low-rank approximation, which aims to enhance the angular resolution of real beam mapping imagery. By constructing a random matrix sketch to sample the raw echo measurements and restore the surface reflectivity map in a low-dimensional linear space, the proposed strategy significantly reduces computational complexity. The Fourier transform-based antenna analysis method is used to determine the optimal low-rank approximation parameter, achieving a balance between error and computational efficiency.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Information Systems
Saba Zia, Atif Qayyum, Fahad Mumtaz Malik, Mohammad Bilal Malik
Summary: This paper presents a control scheme for real-time tracking of nonlinear systems with hard nonlinearities. The proposed scheme introduces a refining component to compensate for performance degradation caused by modelling uncertainties and external disturbances. Simulations on an inverted pendulum demonstrate remarkably superior tracking performance compared to existing techniques.
Article
Thermodynamics
Byoungjoo Chun, S. Mahmood Mousavi, Jongkwon Lee, Bok Jik Lee, Salah A. Faroughi
Summary: This paper investigates the optimization of turbulence in combustion systems using a real-scale low-swirl combustor and different turbulence generation plates. The research reveals that a fractal geometry with a 73% blockage ratio and four iteration levels enhances turbulence intensity and improves combustion efficiency in the nonreacting mode. However, in the reacting mode, a fractal with a 73% blockage ratio and three iteration levels shows the least progress. This study is significant for improving combustion efficiency and reducing emissions.
CASE STUDIES IN THERMAL ENGINEERING
(2023)
Article
Computer Science, Information Systems
Karar Mahmoud, Mohamed Abdel-Nasser, Matti Lehtonen
Summary: The article introduces a low-computational and accurate voltage assessment approach for distribution systems interconnected with photovoltaic (PV) units, which can rapidly compute the voltage deviation in the entire system and terminal voltages of PV units, especially suitable for fine-resolution simulations.
IEEE SYSTEMS JOURNAL
(2021)
Article
Agriculture, Multidisciplinary
Iftach Klapp, Peretz Yafin, Navot Oz, Omri Brand, Idan Bahat, Eitan Goldshtein, Yafit Cohen, Victor Alchanatis, Nir Sochen
Summary: This paper presents progress in overcoming the limitations of low-altitude radiometric aerial surveys for agricultural needs using algorithmic and computational imaging methods. These methods involve stabilizing low-cost thermal cameras to gather radiometric data and enhancing resolution through convolutional neural network-based super-resolution. The combined approach results in a large mosaicked image of the field, providing improved radiometric accuracy and image sharpness.
PRECISION AGRICULTURE
(2021)
Article
Computer Science, Artificial Intelligence
Hao Feng, Liejun Wang, Yongming Li, Anyu Du
Summary: In this paper, a lightweight baseline model LKASR based on large kernel attention (LKA) is proposed for image super-resolution. The model achieves superior performance in image feature extraction and reconstruction by utilizing cascaded visual attention modules and large convolution kernels. Experimental results demonstrate that LKASR outperforms most lightweight methods in different scales of super-resolution.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Mathematics
Young Won Lee, Kang Ryoung Park
Summary: This study investigates high-resolution iris and ocular recognition methods and provides detailed explanations for solving low-resolution issues. Additionally, it introduces the latest deep learning-based approaches.
Article
Engineering, Electrical & Electronic
Shahin Khobahi, Nir Shlezinger, Mojtaba Soltanalian, Yonina C. Eldar
Summary: In this paper, we propose the LoRD-Net, a deep detector designed for recovering information symbols from one-bit noisy measurements. LoRD-Net is a model-aware data-driven architecture based on deep unfolding of optimization iterations, incorporating domain knowledge in its design for data-driven operation with the benefits of model-based optimization methods. Numerical evaluations show that the proposed hybrid methodology of LoRD-Net outperforms existing methods while training on small datasets.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Article
Mathematics, Applied
Gregor Gantner, Alexander Haberl, Dirk Praetorius, Stefan Schimanko
Summary: The paper presents an adaptive finite element method for second-order elliptic PDEs, incorporating an adaptive algorithm that monitors and guides mesh refinement and approximate solutions of discrete systems. It is shown to have linear convergence with optimal algebraic rates, focusing on convergence rates with respect to overall computational costs. Unlike prior works, the proposed adaptive strategy guarantees quasi-optimal computational time, covering both linear problems solved with optimally preconditioned CG method and nonlinear problems linearized by Zarantonello iteration for strongly monotone nonlinearity.
MATHEMATICS OF COMPUTATION
(2021)
Article
Mathematics, Interdisciplinary Applications
Qing Guo, Hongxing Rui
Summary: This paper presents and analyzes two block-centered local refinement methods for solving time-fractional equations, with different approximation methods used for pressure values at slave nodes. The stability analysis is carefully proved and the discrete L-2 error estimates for velocity and pressure are established on locally refinement composite grids. Numerical experiments confirm the convergence rates in agreement with theoretical analysis.
CHAOS SOLITONS & FRACTALS
(2021)
Article
Engineering, Electrical & Electronic
Wenjun Xia, Hongming Shan, Ge Wang, Yi Zhang
Summary: Since 2016, deep learning has made remarkable progress in tomographic imaging, particularly in low-dose computed tomography (LDCT). However, the black-box nature and instabilities of LDCT denoising and end-to-end reconstruction networks hinder the application of DL methods in LDCT. A recent trend is to integrate imaging physics and models into deep networks, allowing for a combination of physics-/model-based and data-driven elements. This article provides a systematic review of physics-/model-based data-driven methods for LDCT, including loss functions, training strategies, performance evaluation, and future directions.
IEEE SIGNAL PROCESSING MAGAZINE
(2023)
Article
Chemistry, Multidisciplinary
Thomas D. Downes, S. Paul Jones, Hanna F. Klein, Mary C. Wheldon, Masakazu Atobe, Paul S. Bond, James D. Firth, Ngai S. Chan, Laura Waddelove, Roderick E. Hubbard, David C. Blakemore, Claudia De Fusco, Stephen D. Roughley, Lewis R. Vidler, Maria Ann Whatton, Alison J. -A. Woolford, Gail L. Wrigley, Peter O'Brien
CHEMISTRY-A EUROPEAN JOURNAL
(2020)
Article
Biochemical Research Methods
Soon Wen Hoh, Tom Burnley, Kevin Cowtan
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2020)
Article
Biochemical Research Methods
Paul S. Bond, Keith S. Wilson, Kevin D. Cowtan
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2020)
Article
Biochemical Research Methods
Emad Alharbi, Radu Calinescu, Kevin Cowtan
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2020)
Letter
Biochemical Research Methods
K. Cowtan
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2020)
Article
Biochemical Research Methods
Catherine L. Lawson, Andriy Kryshtafovych, Paul D. Adams, Pavel Afonine, Matthew L. Baker, Benjamin A. Barad, Paul Bond, Tom Burnley, Renzhi Cao, Jianlin Cheng, Grzegorz Chojnowski, Kevin Cowtan, Ken A. Dill, Frank DiMaio, Daniel P. Farrell, James S. Fraser, Mark A. Herzik, Soon Wen Hoh, Jie Hou, Li-Wei Hung, Maxim Igaev, Agnel P. Joseph, Daisuke Kihara, Dilip Kumar, Sumit Mittal, Bohdan Monastyrskyy, Mateusz Olek, Colin M. Palmer, Ardan Patwardhan, Alberto Perez, Jonas Pfab, Grigore D. Pintilie, Jane S. Richardson, Peter B. Rosenthal, Daipayan Sarkar, Luisa U. Schafer, Michael F. Schmid, Gunnar F. Schroder, Mrinal Shekhar, Dong Si, Abishek Singharoy, Genki Terashi, Thomas C. Terwilliger, Andrea Vaiana, Liguo Wang, Zhe Wang, Stephanie A. Wankowicz, Christopher J. Williams, Martyn Winn, Tianqi Wu, Xiaodi Yu, Kaiming Zhang, Helen M. Berman, Wah Chiu
Summary: The 2019 Cryo-EM Model Challenge evaluated the quality of models produced from cryo-EM maps, reproducibility of modeling results, and performance of metrics used for model validation. The study found relatively high accuracy and reproducibility of cryo-EM models derived by 13 participating teams from four benchmark maps, with recommendations for validating near-atomic cryo-EM structures.
Article
Biochemical Research Methods
Emad Alharbi, Paul Bond, Radu Calinescu, Kevin Cowtan
Summary: Proteins rely on their three-dimensional structure for essential biological functions, but determining this structure through computational work can be challenging. A new software tool has been introduced to accurately predict quality measures of protein structures, aiding researchers in selecting the most effective pipelines for model generation.
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2021)
Article
Biochemical Research Methods
Agnel Praveen Joseph, Mateusz Olek, Sony Malhotra, Peijun Zhang, Kevin Cowtan, Tom Burnley, Martyn D. Winn
Summary: This article introduces a graphical user interface for atomic model validation and discusses the issues of model optimization and refinement. It was found that at low resolutions, stereochemical quality may be favored over data fit, but it is still important to ensure that the model agrees with the data in terms of resolvable features.
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2022)
Article
Biochemical Research Methods
Paul S. Bond, Kevin D. Cowtan
Summary: Interactive model building is a challenging and time-consuming step in the structure-solution process. Automated programs like Buccaneer can simplify and accelerate this process, but they may encounter difficulties with low-resolution data or poor initial models. ModelCraft is a new pipeline that improves model building by expanding on the capabilities of Buccaneer, resulting in increased completeness of protein models in different types of structure solution cases.
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2022)
Article
Biochemical Research Methods
Emad Alharbi, Radu Calinescu, Kevin Cowtan
Summary: A neural network was trained to identify and remove unfavorable fragments in the protein model-building process, resulting in improved backbone tracing. Experimental results showed that using the neural network in Buccaneer software significantly increased the completeness of protein models.
ACTA CRYSTALLOGRAPHICA SECTION D-STRUCTURAL BIOLOGY
(2023)
Article
Meteorology & Atmospheric Sciences
Martin B. Stolpe, Kevin Cowtan, Iselin Medhaug, Reto Knutti
Summary: Climate models have shown discrepancies in simulating global mean temperature changes, especially during the 'global warming hiatus' period in the early twenty-first century. Studies have revealed that the variability in the equatorial Pacific may have played a role in these divergences. While multiple linear regression approaches and pacemaker experiments differ in estimating the Pacific contribution to global temperature during the hiatus, taking into account forced signals in tropical Pacific SST and variations in model sensitivity can help improve the models' performance in reproducing observed global mean temperatures.
