检索结果-内蒙古大学图书馆

Traffic sign recognition model based on scale sequence features and high-order spatial interactions

Neural Computing and Applications 2025年第5期37卷 3605-3620页

作者： Zhang, Gan Wang, Yafei Li, Wenju Fu, Xianping School of Information Science and Technology Dalian Maritime University Dalian116026 China School of Computer Science and Information Engineering Shanghai Institute of Technology Shanghai201418 China Pengcheng Laboratory Shenzhen518055 China

Traffic sign recognition is an integral part of driver assistance systems play a crucial role in enhancing road safety. Due to a large number of challenging targets, such as occlusion, distortion, and small targets in actual scenes, existing methods still have bottlenecks in the accuracy of detection and recognition. Therefore, this paper proposes a real-time traffic sign recognition model based on scale sequence features (SSFs) and high-order spatial interactions (HOSIs) on the basis of YOLOv5. This model extracted scale-invariant SSF using a high-dimensional convolution on the underlying feature map to improve the detection effect of multi-scale targets. At the same time, recursive gated convolution modules are integrated into the feature pyramid network, expanding the interactions to arbitrary order and providing richer feature information for high-order convolutions. In addition, a cross-stage partial (CSP) structure is incorporated into the spatial pooling pyramid, which improves the structure’s performance without introducing excessive computational complexity. The experimental results show that the proposed model achieves 95.5% and 99.92% of mAP on Tsinghua–Tencent 100K (TT100K) and German Traffic Sign Recognition Benchmark (GTSRB), respectively, which surpass the performance of the most advanced model and meets the requirements of real-time detection. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Improvement of Prior Image for Metal Artifact Reduction of Computed Tomography

引用

Journal of Shanghai Jiaotong University (science) 2025年第3期30卷 446-451页

作者： Sun, Wenwu Zhuang, Tiange Chen, Siping School of Computer Science and Engineering Hunan University of Science and Technology Hunan Xiangtan411201 China School of Biomedical Engineering Shanghai Jiao Tong University Shanghai200240 China School of Biomedical Engineering Shenzhen University Guangdong Shenzhen518060 China

It is not easy to reduce the metal artifacts of computed tomography images. However, the pixel values inside the metal artifact regions vary smoothly, while those on the borders of the metal and the bone regions vary sharply. When the Canny operation by adaptive thresholding is conducted on the raw image, the almost continuous edges can be formed obviously on the borders of the metal and the bone regions, but this kind of information cannot be formed for the metal artifact regions. In this paper, by searching the closed areas formed by the border edges of the bone regions in the Canny image, the metal artifact regions, which are very difficult to discriminate only by intensity thresholding, can be excluded effectively. A novel prior image-based method is thus developed for metal artifact reduction. The experiments demonstrate that the proposed method can be realized easily and reduce the metal artifacts effectively even if multiple large metal objects exist simultaneously in the image. The method is suitable for the clinical application. © Shanghai Jiao Tong University 2023.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Multi-domain feature-enhanced attribute updater for generalized zero-shot learning

引用

Neural Computing and Applications 2025年第14期37卷 8397-8414页

作者： Shi, Yuyan Jiang, Chenyi Song, Feifan Ye, Qiaolin Long, Yang Zhang, Haofeng School of Computer Science and Engineering Nanjing University of Science and Technology Jiangsu Nanjing210094 China School of Information Science and Technology Nanjing Forestry University Jiangsu Nanjing210037 China Department of Computer Science Durham University DurhamDH1 3LE United Kingdom

The goal of generalized zero-shot learning (GZSL) is to transfer knowledge from seen classes to unseen classes. However, a significant challenge is the single-category attributes are often inadequate to capture the intra-class variability of diverse image samples. During training, the model tends to cluster samples of seen categories based solely on semantic descriptions, overlooking subtle yet important visual differences. To address these challenges, this paper introduces an instance-level attribute updating method, termed as multi-domain feature-enhanced attribute updater (MDAU), which establishes a one-to-one alignment between visual and semantic by obtaining multiple instance-level attributes that closely correspond to the natural intra-class distribution of the image, in contrast to the many-to-one alignment typically used in traditional GZSL approaches. Specifically, we initially compute feature weights for each individual image, and then, the visual information is incorporated as conditions to facilitate the personalized updating of class attributes. We evaluate our model on three benchmark datasets, the results show that our model can achieve the state-of-the-art performance and detailed analysis also illustrates the effectiveness of each proposed module. Code is available at https://***/S-Silvia/MDAU. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Medical Knowledge Graph:Data Sources,Construction,Reasoning,and Applications

引用

Big Data Mining and Analytics 2023年第2期6卷 201-217页

作者： Xuehong Wu Junwen Duan Yi Pan Min Li Computer Science and Engineering Central South UniversityChangsha 410083China School of Computer Science Hunan First Normal UniversityChangsha 410006China School of Computer Science and Engineering Central South UniversityChangsha 410083China Computer Science and Control Engineering Shenzhen Institute of Advanced TechnologyChinese Academy of SciencesShenzhen 518000China

Medical knowledge graphs(MKGs)are the basis for intelligent health care,and they have been in use in a variety of intelligent medical ***,understanding the research and application development of MKGs will be crucial for future relevant research in the biomedical *** this end,we offer an in-depth review of MKG in this *** research begins with the examination of four types of medical information sources,knowledge graph creation methodologies,and six major themes for MKG ***,three popular models of reasoning from the viewpoint of knowledge reasoning are discussed.A reasoning implementation path(RIP)is proposed as a means of expressing the reasoning procedures for *** addition,we explore intelligent medical applications based on RIP and MKG and classify them into nine major ***,we summarize the current state of MKG research based on more than 130 publications and future challenges and opportunities.

关键词： medical knowledge graph knowledge graph construction knowledge reasoning intelligent medical applications intelligent healthcare

来源：评论

学校读者我要写书评

暂无评论

A METHOD FOR IMPROVING THE ACCURACY OF SFM 3D MEASUREMENTS BY RANDOM PATTERN PROJECTION

引用

AIJ Journal of technology and Design 2023年第71期29卷 533-537页

作者： Sugano, Miki Kakizaki, Ken'ichi Tominaga, Yasumasa Araki, Shunsuke Graduate School of Computer Science and Systems Engineering Kyushu Institute of Technology Japan

SfM can reconstruct the 3D shape of an object with high accuracy if there are many feature points on the surface to be measured. However, in the field of architecture, there is a problem that building surfaces are beautifully preserved and the shape cannot be reconstructed because of the small number of feature points. In this study, we proposed a method to improve the measurement accuracy of SfM by projecting a random pattern on a wall surface and distributing a large number of feature points. Experiments confirmed that the measurement accuracy was as good as that of a laser scanner. © 2023 Architectural Institute of Japan. All rights reserved.

关键词： 3D measurement Measurement accuracy Pattern projection Point cloud Random pattern SfM

来源：评论

学校读者我要写书评

暂无评论

Core-Competitiveness in Partially Observable Networked Market 24

Core-Competitiveness in Partially Observable Networked Marke...

引用

33rd ACM Web Conference, WWW 2024

作者： Li, Bin Hao, Dong School of Computer Science and Engineering Nanjing University of Science and Technology Jiangsu Nanjing China School of Computer Science and Engineering University of Electronic Science and Technology of China Sichuan Chengdu China

ISBN: (纸本)9798400701719

In auction theory, a core is a stable outcome where no subgroup of participants can achieve better results for themselves. Core-competitive auctions aim to generate revenue that is achievable in a core. They are particularly important because they not only generate optimized revenue for the seller, but also provide an efficient and stable environment for participants. We generalize the design of core-competitive auctions to encompass partially observable networked markets (PONM). Unlike traditional auctions, which often deal with scenarios of limited trading activity, our approach to core-competitive auctions for PONM captures the nature of real-world transaction markets, which is a large linking world for the economic entities and commodities circulate among the entities in the market. Our generalizing the auction market to PONM can much improve the liquidity of the auction, and is especially meaningful for the web economics. Specifically, we quantify the upper and lower bounds of the minimum core revenue in PONM, and further prove that there does not exist any truthful auction for PONM which is efficient and core-competitive. Governed by this impossible result, we identify the criteria that the allocation rule for PONM should meet. Based on these criteria, we propose a new class of auction mechanisms for PONM that is individually rational, incentive-compatible, and core-competitive. © 2024 ACM.

关键词： Competition

来源：评论

学校读者我要写书评

暂无评论

Two-dimension high efficiency beam steering based on liquid crystal polarization gratings

引用

Optics Express 2025年第10期33卷 21425-21440页

作者： Li, Zifan Yu, Bijun Gong, Junhao Liu, Yutong Fan, Lina Gao, Xin Yin, Jiahang Ding, Tongshan Zhang, Dawei Wang, Jun Hong, Ruijin Mao, Hongmin Lu, Huanjun Cao, Zhaoliang Key Laboratory of Intelligent Optoelectronic Devices and Chips of Jiangsu Higher Education Institutions School of Physical Science and Technology Suzhou University of Science and Technology Suzhou215009 China Advanced Technology Research Institute of Taihu Photon Center School of Physical Science and Technology Suzhou University of Science and Technology Suzhou215009 China School of Optoelectronic Information and Computer Engineering University of Shanghai for Science and Technology Shanghai200093 China School of Electronics and Information Engineering Suzhou University of Science and Technology Jiangsu Suzhou215009 China

Cascaded liquid crystal polarization grating (LCPG) structure is commonly used to realize two-dimension large angle optical beam steering. In this architecture, different LCPGs have different incident angles. As a liquid crystal half-wave plate (LC HWP) is placed in front of the LCPG to control steering direction, the different incident angles will affect the phase modulation of LC HWP and then, decrease the steering efficiency. In this paper, the steering efficiency of cascaded LCPGs under tilt incidence is considered and a compensation method is proposed to increase the steering efficiency. A simple analysis model is established to calculate the steering efficiency of cascaded LCPG structure for tilt incidence with arbitrary direction. Further, the effect of tilt incidence on the phase modulation of LC HWP is compensated and a simple formula is given to compute the compensation voltage. A validation experiment is performed to verify the proposed method. All the simulated and experimental results indicate that the steering efficiency of cascaded LCPGs can be improved for two-dimension beam scanning. This work is helpful for high efficiency beam steering based on cascaded LCPGs. © 2025 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement.

关键词： Bragg gratings

来源：评论

学校读者我要写书评

暂无评论

A Rearrangement and Restore-Based Mixer Model for Target-Oriented Multimodal Sentiment Classification

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第6期5卷 3109-3119页

作者： Jia, Li Ma, Tinghuai Rong, Huan Sheng, Victor S. Huang, Xuejian Xie, Xintong Nanjing University of Information Science and Technology School of Computer Science Nanjing Jiangsu210044 China Nanjing University of Information Science and Technology School of Computer and Software Nanjing Jiangsu210044 China Jiangsu Ocean University School of Computer Engineering Lianyungang Jiangsu222005 China Nanjing University of Information Science and Technology School of Artificial Intelligence Nanjing Jiangsu210044 China Texas Tech University Department of Computer Science LubbockTX79409 United States

With the development of fine-grained multimodal sentiment analysis tasks, target-oriented multimodal sentiment (TMSC) analysis has received more attention, which aims to classify the sentiment of target with the help of textual and associated image features. Existing methods focus on exploring fine-grained image features and incorporate transformer-based complex fusion strategies, while ignoring the heavy computational burden. Recently, some lightweight multilayer perceptrons (MLP)-based methods have been successfully applied to multimodal sentiment classification tasks. In this article, we propose an effective rearrangement and restore mixer model (RR-Mixer) for TMSC, which dedicates the interaction of image, text, and targets along the modal-axis, sequential-axis, and feature channel-axis through rearrangement and restore operations. Specifically, we take vision transformer (ViT) and robustly optimized BERT (RoBERTa) pretrained models to extract image and textual features, respectively. Further, we adopt cosine similarity to select the most semantically relevant image features. Then, an RR-Mixer module is designed for mixed multimodal features, with the core technology consisting of rolling, grouping rearrangement and restore operations. Moreover, we introduce MLP Unit to learn the information of different modalities for intermodal interaction. The results show that our model achieves superior performance on two benchmark multimodal datasets, TWITTER-15 and TWITTER-17, with a significant improvement of 4.66% and 1.26% in terms of macro-F1. © 2020 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Enhancing Image Super-Resolution with GANs Featuring Enhanced Deep Super-Resolution and Attention Mechanisms 4

Enhancing Image Super-Resolution with GANs Featuring Enhance...

引用

4th IEEE International Conference on Emerging Techniques in Computational Intelligence, ICETCI 2024

作者： Kumar, G. Aditya Kativar, Aditi Kannan, A. Vellore Institute of Technology Computer Science and Engineering Vellore India School of Computer Science and Engineering Vellore Institute of Technology Vellore India

ISBN: (纸本)9798350387063

Improving the quality and resolution of low- resolution digital images is an important task with far-reaching implications for a variety of applications, including medical imaging, surveillance, and content retrieval. To solve the high- resolution problem of a single image, this study proposes a new Generative Adversarial Network (GAN) model that includes an Enhanced Deep Super-Resolution (EDSR) generator. Our model includes a sensitivity function and a Squeeze-and- Excitation (SE) block in an EDSR design. These SE blocks directly model channel connections to optimize channel-specific performance responses and improve network capabilities without incurring significant computational costs. After extensive training using complex data processing methods such as rotation, rotation, and rotation, our model shows an impressive ability to enhance image details and traditional culture more accurately. Our strategy outperforms existing models both statistically and qualitatively with better signal-to- noise ratios (PSNRs) on the DIV2K dataset. These results, together with deep learning models, demonstrate the role of attentional processes in improving image resolution and lay the groundwork for further work in this field. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

User Task Priority Based Resource Allocation with Multi Class Task Scheduling Strategy and Load Balancing in Cloud Environment

引用

SN computer science 2024年第8期5卷 970页

作者： G, Nida Kousar C, Gopala Krishnan Department of Computer Science and Engineering GITAM School of Technology Bengaluru 561203 India

An effective task scheduling method can accommodate user needs, boost resource usage, and boost cloud computing's overall efficiency. However, the unchanging task needs are generally the focus of grid computing's job scheduling, leading to low resource usage. Distributing the dynamic user tasks fairly among all cloud nodes is the goal of load balancing, a relatively new field of study. The primary difficulty with cloud computing is load balancing. By making better use of available resources, load balancing methods improve cloud performance. Load balancing primary goal is to lessen the burden on the environment by cutting down on energy use and carbon emissions. The most crucial characteristics that can both satisfy user needs and maximize resource utilization are used to determine the order of priorities. Existing systems often ignore user priority suggestions in favor of optimal scheduling to improve load balancing. Scheduling that takes into account user-guided priorities uses a data-driven strategy, which helps improve load balancing. Scheduling algorithms that take user priorities into account can optimize load distribution more effectively. The primary objective of this research is to provide a priority based randomized load balancing technique that assigns tasks to virtual machines in a random fashion based on criteria such as the number of users, the amount of time the task takes to run, the type of software being used, the cost of the software, and the amount of available resources. This method maximizes system performance by decreasing response time and resource consumption while increasing metrics like fault tolerance and scalability. This system for scheduling tasks not only accommodates user needs but also achieves excellent resource usage. This research proposes a User Task Priority based Resource Allocation with Multi Class Task Scheduling Strategy and Load Balancing (UPRA-MCTSS-LB) Model for enhancing the cloud service quality. The proposed method res

关键词： Cloud computing Load balancing Multi class tasks Quality of service Resource utilization Task scheduling User task priority

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：