检索结果-内蒙古大学图书馆

6th International Conference on Artificial Intelligence and Pattern Recognition, AIPR 2023

作者： Hua, Wenqiang Sun, Nan Liu, Lin The Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Xi'an Key Laboratory of Big Data and Intelligent Computing School of Computer Science and Technology Xi'an University of Posts and Telecommunications Xi'an China School of Computer Technology Xi'an University of Posts and Telecommunications Xi'an China

ISBN: (纸本)9798400707674

Polarimetric synthetic aperture radar (PolSAR) image classification has important application value and a wide range of application scenarios in many fields. Supervised classification methods, which need to use a large amount of labeled data to improve classification accuracy, have achieved satisfactory classification performance. However, it is difficult and expensive to obtain sufficient labeled data for PolSAR images. The problem to be solved is how to obtain high classification accuracy with a limited number of labeled samples. Semi-supervised PolSAR image classification method based on contrastive learning is proposed. Firstly, a supervised PolSAR image classification method based on meta-learning is introduced. Prototypical network based on meta-learning ideas is used as the basic classification model, and a prototypical representation is created for each category. Classification is performed by computing the similarity between the prototypical representation and samples to be classified. Secondly, a method of training feature extractor based on contrastive learning is designed. By maintaining the consistency of features between the same data and distinguishing the differences between different data after data augmentation, the method enables the feature extractor to obtain accurate feature representation. By co-training with meta-learning, so as to improve the classification performance. Finally, experiments on real PolSAR data prove the effectiveness of the proposed method. © 2023 ACM.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Investigation on the Temperature Stabilization of the Laser Diode Based on PID Control

Investigation on the Temperature Stabilization of the Laser ...

引用

2022 IEEE Delhi Section Conference, DELCON 2022

作者： Tochukwu, Nwodom Eric-Theophilus Tetteh, Akiti Kenneth Agordzo, George K. Zhejiang University of Science and Technology School of Automation Hangzhou China Anhui University of Science and Technology School of Big Data and Mathematics Huainan China

ISBN: (纸本)9781665458832

The most commonly used PI/PID controller is used in industry. Almost all feedback loops in the industry use a P, PI, or PID controller (later PID will be used to refer to all subset controllers). The investigation is designed to stabilize the temperature of laser diodes using PID controller to: (1) stabilize the temperature to a certain extent to prevent sudden shifts;(2) be able to control the temperature using PID algorithms for the successful functioning of temperature-stabilized diodes;and (3) to minimize the fluctuation of temperature. The schematic method is to build a closed-loop system that regulates the laser temperature by using PID algorithms to keep the laser temperature as close to the required set point as possible. The system measures the laser temperature through the thermocouple which sends the acquired temperature to the microcontroller for processing. The system runs with P, PI, and PID controller methods. The Integral gain causes some oscillations but lowers the measurement on the increasing time. The additional integral term to the unit helps lower the steady-state error that is triggered by the P controller, as the integral controller constructs. The rising time is faster than 50 seconds. Many types of PID controllers may be programmed online using machine learning algorithms which improve the efficiency of the controller systems. When it is needed to continuously control a system variable, the PID control is an excellent alternative. © 2022 IEEE.

关键词： Three term control systems

来源：评论

学校读者我要写书评

暂无评论

Autonomous Emergency Landing on 3D Terrains: Approaches for Monocular Vision-based UAVs 8

Autonomous Emergency Landing on 3D Terrains: Approaches for ...

引用

8th IEEE International Conference on Advanced Robotics and Mechatronics, ICARM 2023

作者： Du, Weiming Lin, Junmou Du, Binqing Borhan, Uddin Md. Li, Jianqiang Chen, Jie College of Computer Science and Software Engineering Shenzhen University Shenzhen China Shenzhen University National Engineering Laboratory for Big Data System Computing Technology Shenzhen China

ISBN: (纸本)9798350300178

With the increasing use of unmanned aerial vehicles (UAVs) in a variety of applications, their safety has become a critical concern. UAVs face numerous emergencies during missions, in these situations, the UAVs need to autonomously find a suitable landing site, plan flight routes, and avoid obstacles in unstructured environments. However, due to the limitations of computing power and sensors, it is challenging to achieve this goal. This research aims to explore the monocular emergency autonomous landing algorithm. Specifically, this work focuses on extracting depth and vision information. A topology information extractor is designed to convert image to graph and assess the connectivity of terrain. Additionally, a depth information extractor is designed to calculate the slope and roughness of the ground. A 3D topology optimizer is designed to optimize the graph by depth information and evaluate the landing suitability by a heuristic strategy. To make the action decision, a hybrid decision network (DGN) based on 3D topology information is proposed. Finally, this work build a simulated scenario based on real scene to verify the efficiency of DGN. The results of the experiment show that DGN outperforms its counterparts in terms of action prediction accuracy and landing success rate. © 2023 IEEE.

关键词： Landing

来源：评论

学校读者我要写书评

暂无评论

Deep attention enhanced networks for medical image segmentation

Deep attention enhanced networks for medical image segmentat...

引用

2024 International Conference on Remote Sensing, Mapping, and Image Processing, RSMIP 2024

作者： Shen, Longfeng Wang, Qiong Wang, Wei Wang, Ran Yu, Yexun Zhang, Changchang School of Computer Science and Technology Huaibei Normal University Huaibei China Institute of Artificial lntelligence Hefei Comprehensive National Science Center Hefei China Anhui Big-Data Research Center on University Management Anhui Huaibei China

ISBN: (纸本)9781510680012

Medical image segmentation is a crucial task within the realm of medical image processing. Nevertheless, the intrinsic characteristics of medical images and the limited availability of data constrain the model's generalization capacity. Addressing this challenge requires an infusion of more data and the implementation of effective segmentation techniques to enhance model performance. In response to this need, we propose a deep attention enhanced network for medical image segmentation. This innovative approach boosts the segmentation model's efficacy through techniques such as data augmentation, a deep attention enhanced decoder, and a dual convolutional segmentation head. Validation across multiple datasets substantiates the method's effectiveness and its robust generalization capabilities. © 2024 SPIE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A Method for Identifying the Timeliness of Manufacturing data Based on Weighted Timeliness Graph 9th

A Method for Identifying the Timeliness of Manufacturing Dat...

引用

19th International Conference on Advanced data Mining and Applications, ADMA 2023

作者： Liu, Zehua Ding, Xuefeng Jiang, Yuming Hu, Dasha College of Computer Science Sichuan University Chengdu610065 China Big Data Analysis and Fusion Application Technology Engineering Laboratory of Sichuan Province Chengdu610065 China

ISBN: (纸本)9783031466601

Timeliness is one of the important indicators of data quality. In industrial production processes, a large amount of dependent data is generated, often resulting in unclear timestamps. Therefore, this article combines the conclusion dependency graph into a process dependency graph to determine the identification order of the timeliness of each process data;By constructing a weighted timeliness graph (WTG) and path single flux, a data timeliness identification method that does not completely rely on timestamps is proposed. Finally, a time-effectiveness identification method based on weighted time-effectiveness graph was discussed through an example and 9 dependency rules, and the effectiveness of the method was verified through a set of experiments. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： big data

来源：评论

学校读者我要写书评

暂无评论

Cycle Based Link Prediction Method for Cross-Granularity Contrast Learning

Cycle Based Link Prediction Method for Cross-Granularity Con...

引用

Medical Artificial Intelligence (MedAI), IEEE International Conference on

作者： Zehua Zhang Feihu Zheng Bin Wang Surui Yan Faculty of Computer Science and Technology (School of Big Data) Taiyuan University of Technology Jinzhong China

ISBN: (数字)9798350377613

ISBN: (纸本)9798350377620

In the process of network evolution, interactions be-tween network nodes often extend beyond simple binary interactions to include group interactions. Examples include social networks like those found on Facebook and research collaboration networks formed through scientific partnerships. Link prediction in research collaboration networks aims to forecast collaborative relationships among scientists, thereby enhancing our understanding and opti-mizing the structure and dynamics of these networks. However, net-work data are frequently sparse, and prediction based solely on at-tributes or structural features often yields inconsistent results. In complex networks, higher-order structures often contain more comprehensive information. Capturing the cycle structure in the graph enriches the representation of nodes as well as reduces the infor-mation loss during the information transfer process, thus improving the accuracy of link prediction between nodes. Specifically, our approach utilizes the cycle to calculate its significance and reconstructs ‘fat node’ information, which replaces node representations within the original cycle. We introduce a cycle scale module to de-termine the optimal cycle size for the graph. Moreover, by constructing the clustering coefficients of the multivariate cycle, we identify node and cluster aggregations within the graph, thereby reduce the complexity of finding cycle scales. Subsequently, a graph compari-son learning model is employed to establish a mapping between the newly generated ‘fat nodes’ and link relationships. Through experi-mental validation, our method demonstrates strong link prediction performance across various real network datasets.

关键词： Accuracy Social networking (online) Computational modeling Collaboration Predictive models Prediction algorithms Complexity theory Computational efficiency Convolutional neural networks Optimization

来源：评论

学校读者我要写书评

暂无评论

Leveraging Foundation Models for Crafting Narrative Visualization: A Survey

引用

IEEE Transactions on Visualization and Computer Graphics 2025年 PP卷 PP页

作者： He, Yi Xu, Ke Cao, Shixiong Shi, Yang Chen, Qing Cao, Nan Tongji University Intelligent Big Data Visualization Lab China Nanjing University School of Intelligence Science and Technology China

Narrative visualization transforms data into engaging stories, making complex information accessible to a broad audience. Foundation models, with their advanced capabilities such as natural language processing, content generation, and multimodal integration, hold substantial potential for enriching narrative visualization. Recently, a collection of techniques have been introduced for crafting narrative visualizations based on foundation models from different aspects. We build our survey upon 66 papers to study how foundation models can progressively engage in this process and then propose a reference model categorizing the reviewed literature into four essential phases: Analysis, Narration, Visualization, and Interaction. Furthermore, we identify eight specific tasks (e.g. Insight Extraction and Authoring) where foundation models are applied across these stages to facilitate the creation of visual narratives. Detailed descriptions, related literature, and reflections are presented for each task. To make it a more impactful and informative experience for diverse readers, we discuss key research problems and provide the strengths and weaknesses in each task to guide people in identifying and seizing opportunities while navigating challenges in this field. © 2025 IEEE.

关键词： Visualization

来源：评论

学校读者我要写书评

暂无评论

Knowledge Graph based Mutual Attention for Machine Reading Comprehension over Anti-Terrorism Corpus

引用

data Intelligence 2023年第3期5卷 685-706页

作者： Feng Gao Jin Hou Jinguang Gu Lihua Zhang School of Computer Science and Technology Wuhan University of Science and TechnologyWuhan 430065Hubei The Key Laboratory of Rich-Media Knowledge Organization and Service of Digital Publishing Content Insitute of Scientic and Technical Information of ChinaBeijing 100038China Wuhan University of Science and Technology Big Data Science and Engineering Research Institute Wuhan 430065Hubei Eastchina Jiaotong University Nanchang 330013Jiangxi

Machine reading comprehension has been a research focus in natural language processing and intelligence ***,there is a lack of models and datasets for the MRC tasks in the anti-terrorism ***,current research lacks the ability to embed accurate background knowledge and provide precise *** address these two problems,this paper first builds a text corpus and testbed that focuses on the anti-terrorism domain in a semi-automatic ***,it proposes a knowledge-based machine reading comprehension model that fuses domain-related triples from a large-scale encyclopedic knowledge base to enhance the semantics of the *** eliminate knowledge noise that could lead to semantic deviation,this paper uses a mixed mutual ttention mechanism among questions,passages,and knowledge triples to select the most relevant triples before embedding their semantics into the *** results indicate that the proposed approach can achieve a 70.70%EM value and an 87.91%F1 score,with a 4.23%and 3.35%improvement over existing methods,respectively.

关键词： Machine reading comprehension Anti-terrorism domain Knowledge embedding Knowledge attention Mutual attention

来源：评论

学校读者我要写书评

暂无评论

Dynamical relaxation of a long-range Kitaev chain

引用

Physical Review B 2024年第9期109卷 094309-094309页

作者： Yu-Huan Huang Yin-Tao Zou Chengxiang Ding School of Microelectronics and Data Science Anhui University of Technology Maanshan 243002 China Anhui Provincial Joint Key Laboratory of Disciplines for Industrial Big Data Analysis and Intelligent Decision Maanshan 243002 China

We study the universal real-time relaxation behaviors of a long-range Kitaev chain following a quench, and the results are compared to a short-range quantum XY chain. Our research includes both noncritical and critical quenches. In the case of noncritical quench, i.e., neither the initial state nor the postquench Hamiltonian is at a critical point of the equilibrium phase transition, a quench to the commensurate phase or incommensurate phase gives a scaling of t−3/2 or t−1/2, respectively, which is the same as the counterpart of the short-range quantum XY model. However, for a quench to the boundary line between the commensurate and incommensurate phases, the scaling law t−μ may be different from the t−3/4 law of the counterpart of the short-range quantum XY model. More interestingly, the decaying exponent μ may depend on the choice of the parameters of the postquench Hamiltonian because of the different asymptotic behaviors of the energy spectrum. Furthermore, in certain cases, the scaling behavior may be outside the range of predictions made by the stationary phase approximation, because an inflection point emerges in the energy spectrum. For the critical quench, i.e., the initial state or the postquench Hamiltonian is at a critical point of equilibrium phase transition, the aforementioned scaling law t−μ may be changed because of the gap-closing property of the energy spectrum of the critical point.

关键词： Quantum quench Spin relaxation Exact solutions for many-body systems Integrable systems

来源：评论

学校读者我要写书评

暂无评论

Unveiling and Mitigating Backdoor Vulnerabilities based on Unlearning Weight Changes and Backdoor Activeness 38

Unveiling and Mitigating Backdoor Vulnerabilities based on U...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Lin, Weilin Liu, Li Wei, Shaokui Li, Jianze Xiong, Hui The Hong Kong University of Science and Technology Guangzhou China The Chinese University of Hong Kong Shenzhen China Shenzhen International Center for Industrial and Applied Mathematics China Shenzhen Research Institute of Big Data China

The security threat of backdoor attacks is a central concern for deep neural networks (DNNs). Recently, without poisoned data, unlearning models with clean data and then learning a pruning mask have contributed to backdoor defense. Additionally, vanilla fine-tuning with those clean data can help recover the lost clean accuracy. However, the behavior of clean unlearning is still under-explored, and vanilla fine-tuning unintentionally induces back the backdoor effect. In this work, we first investigate model unlearning from the perspective of weight changes and gradient norms, and find two interesting observations in the backdoored model: 1) the weight changes between poison and clean unlearning are positively correlated, making it possible for us to identify the backdoored-related neurons without using poisoned data;2) the neurons of the backdoored model are more active (i.e., larger gradient norm) than those in the clean model, suggesting the need to suppress the gradient norm during fine-tuning. Then, we propose an effective two-stage defense method. In the first stage, an efficient Neuron Weight Change (NWC)-based Backdoor Reinitialization is proposed based on observation 1). In the second stage, based on observation 2), we design an Activeness-Aware Fine-Tuning to replace the vanilla fine-tuning. Extensive experiments, involving eight backdoor attacks on three benchmark datasets, demonstrate the superior performance of our proposed method compared to recent state-of-the-art backdoor defense approaches. The code is available at https://***/linweiii/***. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

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：