检索结果-内蒙古大学图书馆

Neuromorphic Vision-Based Motion Segmentation With Graph Transformer Neural Network

IEEE TRANSACTIONS ON MULTIMEDIA 2025年 27卷 385-400页

作者： Alkendi, Yusra Azzam, Rana Javed, Sajid Seneviratne, Lakmal Zweiri, Yahya Technol Innovat Inst TII Prop & Space Res Ctr PSRC Abu Dhabi U Arab Emirates Khalifa Univ Sci & Technol Khalifa Univ Ctr Autonomous Robot Syst KUCARS Abu Dhabi U Arab Emirates Khalifa Univ Sci & Technol Dept Aerosp Engn Abu Dhabi U Arab Emirates Khalifa Univ Sci & Technol Adv Res & Innovat Ctr ARIC Abu Dhabi U Arab Emirates

Moving object segmentation is critical to interpret scene dynamics for robotic navigation systems in challenging environments. Neuromorphic vision sensors are tailored for motion perception due to their asynchronous nature, high temporal resolution, and reduced power consumption. However, their unconventional output requires novel perception paradigms to leverage their spatially sparse and temporally dense nature. In this work, we propose a novel event-based motion segmentation algorithm using a Graph Transformer Neural Network, dubbed GTNN. Our proposed algorithm processes event streams as 3D graphs by a series of nonlinear transformations to unveil local and global spatiotemporal correlations between events. Based on these correlations, events belonging to moving objects are segmented from the background without prior knowledge of the dynamic scene geometry. The algorithm is trained on publicly available datasets including MOD, EV-IMO, and EV-IMO2 using the proposed training scheme to facilitate efficient training on extensive datasets. Moreover, we introduce the Dynamic Object Mask-aware Event Labeling (DOMEL) approach for generating approximate ground-truth labels for event-based motion segmentation datasets. We use DOMEL to label our own recorded Event dataset for Motion Segmentation (EMS-DOMEL), which we release to the public for further research and benchmarking. Rigorous experiments are conducted on several unseen publicly-available datasets where the results revealed that GTNN outperforms state-of-the-art methods in the presence of dynamic background variations, motion patterns, and multiple dynamic objects with varying sizes and velocities. GTNN achieves significant performance gains with an average increase of 9.4% and 4.5% in terms of motion segmentation accuracy (IoU%) and detection rate (DR%), respectively.

关键词： Motion segmentation Computer vision Dynamics Cameras Event detection Transformers Streams Heuristic algorithms Vehicle dynamics classification algorithms Neuromorphic vision dynamic vision sensor event camera motion segmentation graph transformer neural networks

来源：评论

学校读者我要写书评

暂无评论

Development and Application of Coverage Control algorithms: A Concise Review

引用

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING 2025年 22卷 14906-14927页

作者： Cheng, Bin He, Mingyuan Zhu, Zhongpan He, Bin Chen, Jie Tongji Univ Dept Control Sci & Engn Shanghai 201804 Peoples R China Natl Key Lab Autonomous Intelligent Unmanned Syst Shanghai 201203 Peoples R China

Coverage control is a foundational domain within multi-agent systems, which has recently undergone substantial advancements. Model-based and optimization-based control methods have achieved new breakthroughs and applications, while data-driven and learning-based algorithms in coverage control have also yielded significant results. This review examines the evolution and application of coverage control algorithms in multi-agent systems. Moreover, this review then focuses on how contemporary algorithms build on traditional control strategies and integrate cutting-edge data-driven and adaptive learning techniques. Finally, it explores potential future directions, emphasizing the importance of interdisciplinary approaches in overcoming existing challenges and seizing new opportunities in coverage control deployment. This review aims to be a valuable resource for researchers and practitioners, guiding continued exploration in this field. Note to Practitioners-This work was motivated by the need to provide an overview of the development and research of existing multi-agent coverage control algorithms. Different application scenarios and constraints require different methods of multi-agent coverage control, each with varying efficiencies. For researchers and practitioners, deciding which coverage control method to use in specific scenarios, while considering practical constraints such as energy limitations, can be challenging. Therefore, this work offers an overview of existing multi-agent coverage control algorithms and suggests potential application scenarios. This review serves as a comprehensive reference for researchers and practitioners in the field, helping them identify suitable solutions to ensure efficient coverage.

关键词： Sensors Monitoring Wireless sensor networks Reviews classification algorithms Heuristic algorithms Multi-agent systems Intelligent sensors Collaboration Optimization Coverage control multi-agent systems Voronoi partition multi-agent reinforcement learning wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

Binocular Vision-Based Target Detection Algorithm

引用

IEEE ACCESS 2025年 13卷 73440-73450页

作者： Zhang, Huiguo Xian Univ Architecture & Technol Xian 710055 Shaanxi Peoples R China

In the field of target detection, algorithms are challenged with multi-objective optimization problems in identifying detection targets, and it is also crucial to improve the recognition of small and insignificant targets. In this study, the Feffol network model is proposed to achieve excellent performance in complex detection tasks. In the feature extraction stage, the Feffol network model carefully chooses Efficient-v2 as the backbone network structure, which can provide a solid foundation for subsequent detection by taking advantage of its efficient feature extraction capability. In addition, an Ebifpn feature pyramid module with an spatial pyramid pooling-cross-stage partial connections(SppCSP) structure was innovatively introduced. This design not only effectively expands the feature sensing field, but also greatly enhances the fusion and expression of feature information of different sizes, so that the model can capture the target features more comprehensively. In addition, to solve the key problems in the detection process, the Feffol model adopts the Focal Loss and CIoU Loss functions. The former can effectively balance the positive and negative samples to avoid over-learning of easy-to-classify samples during the training process;the latter successfully solves the problem of the failure of the detection method when there is no intersection between the prediction frame and the real detection frame, which significantly improves the robustness of the model. The experimental results show that the Feffol network model has excellent performance, with a detection accuracy as high as 90.09% and a detection speed of 9.75it/s. Compared with other mainstream networks, the detection accuracy of Feffol network has been significantly improved, which fully proves its advancement and effectiveness in the task of target detection, and provides new ideas and methods for the further development of this field.

关键词： Feature extraction Object detection Accuracy Convolution Target recognition Training Attention mechanisms Laser radar classification algorithms Radar detection Targeted detection complex environmental tasks highly robust Feffol

来源：评论

学校读者我要写书评

暂无评论

RelHDx: Hyperdimensional Computing for Learning on Graphs With FeFET Acceleration

引用

IEEE TRANSACTIONS ON COMPUTERS 2025年第5期74卷 1730-1742页

作者： Kang, Jaeyoung Zhou, Minxuan Xu, Weihong Rosing, Tajana Univ Calif San Diego La Jolla CA 92093 USA

Graph neural networks (GNNs) are a powerful machine learning (ML) method to analyze graph data. The training of GNN has compute and memory-intensive phases along with irregular data movements, which makes in-memory acceleration challenging. We present a hyperdimensional computing (HDC)-based graph ML framework called RelHDx that aggregates node features and graph structure, along with representing node and edge information in high-dimensional space. RelHDx enables single-pass training and inference with simple arithmetic operations, resulting in the efficient design of graph-based ML tasks: node classification and link prediction. We accelerate RelHDx using scalable processing in-memory (PIM) architecture based on emerging ferroelectric FET (FeFET) technology. Our accelerator uses a data allocation optimization and operation scheduler to address the irregularity of the graph and maximize the performance. Evaluation results show that RelHDx offers comparable accuracy to popular GNN-based algorithms while achieving up to $63.8\boldsymbol{\times}$63.8x faster speed on GPU. Our FeFET-based accelerator, RelHDx-PIM, is $32\boldsymbol{\times}$32x faster for node classification, while for link prediction it is $65.4\boldsymbol{\times}$65.4x faster than when running on GPU. Furthermore, RelHDx-PIM improves energy efficiency by four orders of magnitude over GPU. Compared to the state-of-the-art in-memory processing-based GNN accelerator, PIM-GCN [1], RelHDx-PIM is $10\boldsymbol{\times}$10x faster and $986\boldsymbol{\times}$986x more energy-efficient on average.

关键词： FeFETs Training Energy efficiency Encoding Graphics processing units Logic gates Accuracy Vectors classification algorithms Graph neural networks Hyperdimensional computing graph-based machine learning graph neural networks processing-in-memory ferroelectric field-effect transistor

来源：评论

学校读者我要写书评

暂无评论

Land Use Mapping of the Guangdong-Hong Kong Macao Greater Bay Area Based on a New Approach at 30 m Resolution for the Years 1976 to 2020

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2025年 18卷 3943-3958页

作者： Gu, Yu Chen, Yangbo Liu, Jun Sun Yat Sen Univ Sch Geog & Planning Guangzhou 510275 Peoples R China

Multicategory land use data of high spatiotemporal resolution and large scale are crucial for studying regional ecological and environmental changes and urbanization impacts as well as for sustainable development planning. Currently available public data products include those of high spatial resolution global land use temporally limited to a single or short period, or global annual land cover products in which only a single land use type is depicted, such that regional characteristics are overlooked. In either case, fine-scale annual variation over longer time spans may not be reflected. In this study, the Google Earth Engine platform, Landsat satellite imagery, and a substantial number of manually interpreted samples were used to develop a dataset of annual land use changes in the Guangdong-Hong Kong Macao Greater Bay Area (GBA) at a 30 m resolution for the years 1976 to 2020. This dataset, termed Annual Land Use/Cover of the Greater Bay Area (LUC-GBA), was used to analyze the annual land use variation in 11 cities within the GBA. The high level of accuracy achieved with the LUC-GBA dataset was evidenced by an overall accuracy (OA) of 93.9% in 2020. The OA of interannual classification models ranged from 83.9% to 93.9%, and the kappa coefficients from 0.805 to 0.923. These results indicate that the LUC-GBA dataset effectively reflects the surface cover distribution and interannual dynamic evolution of the land area in the GBA at a 30 m spatial resolution, thus providing reliable data support for land surface process research and related applications.

关键词： Urban areas Remote sensing Landsat Land surface Spatial resolution Earth Sensors classification algorithms Satellite images Planning 30-meter resolution annual land use change GBA landsat satellite imagery remote sensing

来源：评论

学校读者我要写书评

暂无评论

TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model

引用

IEEE ACCESS 2025年 13卷 852-865页

作者： Zhang, Zhixiang Liu, Tong Gao, Jinyu Yang, Meng Luo, Wenjun Lin, Fanqiang Chengdu Univ Technol Coll Mech & Elect Engn Chengdu 610059 Peoples R China Santai Cty Educ & Sports Bur Mianyang 621100 Peoples R China

Currently, deep learning applications in tomato disease recognition are becoming widespread. However, in greenhouse tomato cultivation environments, many techniques neglect the foggy image caused by wet conditions, adversely affecting the performance of disease recognition models, which still remains unresolved. To alleviate adverse environmental impacts, such as fog, and enhance the precision and robustness of the model, this study introduces an improved dark channel prior (DCP) dehazing algorithm and proposes the TDR-Model based on the MobileNetV3. Employing the quad-tree spatial hierarchy search strategy, guided filtering, tolerance mechansim, and multi-scale color restoration (MSRCR) algorithm, an innovative improved version of DCP dehazing algorithm is realized. Image preprocessing with the improved DCP dehazing algorithm obtains more authentic dehazed images. Moreover, MobileNetV3 is improved by the incorporation of the convolutional block attention module (CBAM) and Omni-Dimensional Dynamic Convolution(ODC), improving the accuracy of disease feature recognition. The experiment results demonstrate that, with the parameter size remaining nearly unchanged (a slight reduction of 0.04M), an increase of 16.88% is achieved in accuracy compared to the original version for foggy images. Additionally, we provide extensive ablation studies to evaluate effectiveness of the proposed framework.

关键词： Diseases Image recognition Accuracy classification algorithms Atmospheric modeling Humidity Filtering Deep learning Crops Symbols Tomato disease recognition dark channel prior dehazing algorithm deep learning image preprocessing lightweight model

来源：评论

学校读者我要写书评

暂无评论

Adaptive Feature-Manipulated Vehicle and Pedestrian Detection in Infrared Images

引用

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS 2025年第4期26卷 4579-4591页

作者： Chen, Guangchen Zhang, Pengcheng Zhang, Yinhui He, Zifen Shi, Benjie Kunming Univ Sci & Technol Fac Mech & Elect Engn Kunming 650500 Peoples R China

Infrared imaging is widely applied in assisted driving systems to enable night vision for sensitive targets such as vehicles and pedestrians. However, the detection accuracy of these targets is limited on lightweight detection networks due to their coarse color and texture characteristics rendered in infrared images. To solve this problem, we propose an adaptive feature-manipulated network (AFMNet) for accurate vehicle and pedestrian detection in infrared images. First, a refined spatial pooling module that uses one-dimensional convolution is proposed to establish local and channel feature mapping under different receptive fields so that different fine-grained features are fused. Second, the shuffle manipulation module is designed which includes slicing and shuffling to manipulate the spatial and channel features so that the information loss problem caused by conventional convolutional downsampling is overcome. Third, the adaptive connection of features at different scales using learnable parameters is proposed, and then the target features are reinforced by location and channel calibration branches. The experimental results show that AFMNet achieves the best performance in terms of average detection accuracy of 86.4%, model size of 5.3MB, and detection speed of 48FPS on GTX 2080Ti.

关键词： Feature extraction Pedestrians Accuracy Prediction algorithms Object detection Convolutional neural networks Semantics classification algorithms Deep learning Radio frequency Assisted driving infrared detection refined spatial pooling shuffle manipulation adaptive connection

来源：评论

学校读者我要写书评

暂无评论

A Label Propagation Approach for Missing Data Imputation

引用

IEEE ACCESS 2025年 13卷 65925-65938页

作者： Loyola Lopes, Filipe Dantas Mangussi, Arthur Cardoso Pereira, Ricardo Seoane Santos, Miriam Henriques Abreu, Pedro Carolina Lorena, Ana Aeronaut Inst Technol ITA Comp Sci Div BR-12228900 Sao Jose Dos Campos Brazil Fed Univ Sao Paulo UNIFESP Sci & Technol Inst BR-12247014 Sao Jose Dos Campos Brazil Ctr Informat Technol Renato Archer CTI Renato Arch Comp Methodol Div BR-13069901 Campinas Brazil Univ Coimbra Ctr Informat & Syst Univ Coimbra Fac Sci & Technol Coimbra Dept Informat Engn P-3030790 Coimbra Portugal Univ Porto Dept Comp Sci P-4099002 Porto Portugal Inst Syst & Comp Engn Technol & Sci INESC TEC LIAAD P-4200465 Porto Portugal

Missing data is a common challenge in real-world datasets and can arise for various reasons. This has led to the classification of missing data mechanisms as missing completely at random, missing at random, or missing not at random. Currently, the literature offers various algorithms for imputing missing data, each with advantages tailored to specific mechanisms and levels of missingness. This paper introduces a novel approach to missing data imputation using the well-established label propagation algorithm, named Label Propagation for Missing Data Imputation (LPMD). The method combines, weighs, and propagates known feature values to impute missing data. Experiments on benchmark datasets highlight its effectiveness across various missing data scenarios, demonstrating more stable results compared to baseline methods under different missingness mechanisms and levels. The algorithms were evaluated based on processing time, imputation quality (measured by mean absolute error), and impact on classification performance. A variant of the algorithm (LPMD2) generally achieved the fastest processing time compared to other five imputation algorithms from the literature, with speed-ups ranging from 0.7 to 23 times. The results of LPMD were also stable regarding the mean absolute error of the imputed values compared to their original counterparts, for different missing data mechanisms and rates of missing values. In real applications, missingness can behave according to different and unknown mechanisms, so an imputation algorithm that behaves stably for different mechanisms is advantageous. The results regarding ML models produced using the imputed datasets were also comparable to the baselines.

关键词： Imputation Nearest neighbor methods Amputation Data models classification algorithms Autoencoders Support vector machines Statistical analysis Semisupervised learning Self-organizing feature maps Label propagation missing data data imputation data amputation

来源：评论

学校读者我要写书评

暂无评论

A Special Points and Neural Network-Based Dynamic Multi-Objective Optimization Algorithm

引用

IEEE ACCESS 2025年 13卷 24765-24792页

作者： Li, Sanyi Hou, Wenjie Liu, Peng Yue, Weichao Wang, Qian Zhengzhou Univ Light Ind Sch Elect & Informat Engn Zhengzhou 450002 Peoples R China

This paper introduces a special points and neural network- based dynamic multi-objective optimization algorithm (SPNN-DMOA) for solving dynamic multi-objective optimization problems (DMOPs) with an irregularly changing pareto set. In the stage of population initialization, the algorithm employs a feedforward neural network (FNN) along with special points to generate an initial population. The FNN is trained with historical special points (knee point, boundary point, center point), and the current special points are generated by the FNN when an environmental change is detected. Then the decision variables are classified into convergence variables and diversity variables. The convergence variables of special points are locally searched to form a new population and the best individuals of this population are selected. Finally, a portion of the initial population is generated by conducting a local search on the diversity variables of best individuals, while the remaining portion is produced using random strategies. SPNN-DMOA only needs to maintain non-dominated solutions in proximity to special points, which reduces the computational complexity in the dynamic evolution process. The proposed algorithm has been compared with other state-of-the-art algorithms on a series of benchmark problems, demonstrating its superior performance in optimizing DMOPs.

关键词： Heuristic algorithms Optimization Convergence Pareto optimization classification algorithms Prediction algorithms Diversity reception Vectors Optical fibers Machine learning algorithms Dynamic multi-objective optimization decision variable classification irregular environment neural network prediction special point

来源：评论

学校读者我要写书评

暂无评论

Deep Minimax Classifiers for Imbalanced Datasets With a Small Number of Minority Samples

引用

IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING 2025年第3期19卷 491-506页

作者： Choi, Hansung Seo, Daewon DGIST Elect Engn & Comp Sci Daegu 42988 South Korea

The concept of a minimax classifier is well-established in statistical decision theory, but its implementation via neural networks remains challenging, particularly in scenarios with imbalanced training data having a limited number of samples for minority classes. To address this issue, we propose a novel minimax learning algorithm designed to minimize the risk of worst-performing classes. Our algorithm iterates through two steps: a minimization step that trains the model based on a selected target prior, and a maximization step that updates the target prior towards the adversarial prior for the trained model. In the minimization, we introduce a targeted logit-adjustment loss function that efficiently identifies optimal decision boundaries under the target prior. Moreover, based on a new prior-dependent generalization bound that we obtained, we theoretically prove that our loss function has a better generalization capability than existing loss functions. During the maximization, we refine the target prior by shifting it towards the adversarial prior, depending on the worst-performing classes rather than on per-class risk estimates. Our maximization method is particularly robust in the regime of a small number of samples. Additionally, to adapt to overparameterized neural networks, we partition the entire training dataset into two subsets: one for model training during the minimization step and the other for updating the target prior during the maximization step. Our proposed algorithm has a provable convergence property, and empirical results indicate that our algorithm performs better than or is comparable to existing methods.

关键词： Training Training data Minimization Accuracy Signal processing algorithms Neural networks classification algorithms Feature extraction Data models Decision theory Minimax training imbalanced data adversarial prior

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：